The Inference Report

Automated Text-to-Speech (TTS) evaluation methods (Mean Opinion Score (MOS) predictors and Audio Large Language Models (Audio-LLM) judges) are expected to reflect human perception, yet it is unclear how well they capture the distinct aspects of speech that listeners actually perceive. We deconstruct "naturalness" into a linguistically grounded annotation schema spanning 10 distinct perceptual dimensions, and use it to construct the first dimension-level meta-evaluation benchmark for TTS, comprising 860 utterances annotated by trained linguist raters. Results from benchmarking four MOS predictors and four Audio-LLM judges reveal that MOS predictors collapse onto acoustic signal quality, while Audio-LLM judges show selective, prompt-dependent detection that does not generalise across all dimensions. Neither class reliably captures a breadth of linguistically structured speech errors. Our dataset, annotation schema, and evaluation code are publicly released to support more targeted and interpretable TTS evaluation.

Multimodal Large Language Models (MLLMs) exhibit strong visual understanding, yet the internal features that cause these behaviors remain difficult to identify, audit, or control. While applicable to post-hoc inspection, hidden states that are decomposed into interpretable feature directions using sparse autoencoders (SAEs) neither readily isolate which features are changed by multimodal training, nor are they directly useful for targeted control. We introduce MMDiff, a multimodal model-diffing framework that trains multimodal SAEs and turns them into feature-level interfaces for discovering and controlling multimodal behavior. MMDiff supports three uses: (i) feature isolation, by diffing a base-LM SAE against its multimodal-adapted counterpart to identify features altered by multimodal training; (ii) task-specific feature detection, via per-token contrastive firing analysis that isolates causal features; and (iii) feature-level control, by causally removing or steering the discovered feature directions. We train multimodal SAEs for three MLLM families, LLaVA-MORE, PaliGemma 2, and InternVL3.5, and evaluate on visual-spatial understanding, multimodal safety, and OCR. MMDiff discovers sparse, causally specific features whose removal selectively degrades target behaviors by an average of 12% on spatial tasks and 17% on OCR, and reduces attack success rate by 24% on multimodal safety attacks, with no impact on VQA performance. Steering these features improves spatial and OCR accuracy by +3.6% and +1.8% on average over a standard single-layer steering baseline. These results show that multimodal SAEs can serve not only as interpretability tools, but as mechanisms for auditing, steering, and controlling MLLMs behavior toward safer and more capable generations.

Large language models are increasingly being deployed in governmental settings, yet few existing evaluation frameworks jointly reflect the values of public administration and the linguistic requirements of non-English contexts. We present the "Grip on LLMs" framework, a systematic evaluation suite for Dutch governmental use developed in collaboration with domain experts from a major Dutch municipal organisation. Through an advisory board process, user research, and a survey of the users of a civil-servant chatbot, we identify six evaluation dimensions (factuality, honesty, social bias, energy consumption, cost, and training data transparency) and operationalise them into a benchmark suite covering more than 30 multilingual and Dutch-specific models. Our results reveal that no single model excels across all dimensions, and that trade-offs are unavoidable: higher quality consistently comes at greater environmental impact and financial cost, while bias remains largely independent of both. We further find that factuality (whether a model answers correctly) and honesty (whether a model acknowledges what it does not know) are governed by distinct properties, with high factuality not implying high honesty. To make these findings actionable for non-technical audiences, we release a publicly accessible, user-friendly model overview designed for the full range of stakeholders involved in governmental LLM selection, from engineers to policymakers.

Foundation models are transforming business workflows and boosting productivity, yet they remain largely absent from engineering domains such as power system analysis, where strict physical consistency must be enforced. We present GENCO (GEometric Neural Corrective Optimizer), a unified neural solver for steady-state transmission grid analysis that handles power flow (PF), optimal power flow (OPF), and state estimation (SE) within a single architecture and shared network representation. To support advances in neural power system solvers, we introduce the open-source GridFM Development Framework, which standardizes synthetic data generation and training in a low-code environment. We also release large-scale datasets with millions of PF and OPF scenarios across diverse grid topologies to support reproducible benchmarking. We evaluate GENCO on the PFDelta and OPFData benchmarks against state-of-the-art neural solvers and classical solvers, including Newton-Raphson and IPOPT, as well as on real-world Hydro-Québec SCADA data. For large-scale PF, GENCO recovers the full AC operating state, including voltage magnitudes and reactive power that DC-PF cannot provide, while matching DC-PF-level active power-balance residuals. It achieves up to 30x speedups over Newton-Raphson at only 2x the runtime of DC-PF. For OPF, it achieves up to 85x speedups over IPOPT while improving feasibility, optimality, and runtime over DC-OPF. For SE, GENCO is more robust than classical weighted least squares to noisy measurements and network parameter errors, and always returns a high-quality estimate even when weighted least squares fails to converge. Together, the unified architecture and development framework provide a new approach to large-scale steady-state grid analysis, lowering the barrier to entry for power system engineers and marking a step toward Grid Foundation Models.

Python is a widely adopted programming language, valued for its simplicity and flexibility. However, automated refactoring for Python remains challenging, even though refactoring is an essential practice in software evolution aimed at improving internal code structure without changing external behavior. Understanding how behavioral changes are introduced during refactoring is crucial, as such issues can compromise software reliability and reduce developer productivity. We propose an approach based on a foundation model oracle that analyzes git-style diffs to identify behavioral changes introduced by Python refactorings. We evaluated our technique on Rope refactoring implementations, reusing 1,152 refactoring attempts from a prior study and analyzing 217 resulting transformation pairs with the oracle. Our model-based analysis uncovered 13 distinct bugs among the seven refactoring types studied. All reported bugs were submitted to the respective developers, and 12 of the 13 resulting issue reports were accepted according to issue-tracker evidence. These results highlight the need to improve the robustness of current Python refactoring tools to ensure the correctness of automated code transformations and support reliable software maintenance.

We introduce the Dark Souls Learning Environment (DSLE), a containerized platform that presents all 22 boss encounters of Dark Souls: Remastered as game-playing agent benchmarks through a Gymnasium-style interface. DSLE combines real-time combat, high-dimensional visual input, and sparse terminal rewards, with each environment step being a real action executed against the running game. To support controlled comparison, we define DSLE-5, a representative five-boss subset, spanning a melee fight, a spatially constrained arena, an environmental-hazard fight, a multi-target fight, and a fast final-boss fight, that we recommend as the starting suite for agents built on DSLE. On DSLE-5 we evaluate a random policy, an expert system, an evolutionary baseline, and PPO and DQN agents trained from visual input. The expert system and the evolutionary baseline each defeat the Asylum Demon, the game's tutorial boss (63% and 43% peak win rates), but none of the five methods defeats the other four DSLE-5 bosses; PPO and DQN show no measurable learning (at most 0.33% win rate on the tutorial boss, 0% elsewhere) within a budget that already costs tens of wall-clock hours per run. A broader study running the evolutionary baseline across all 22 encounters under advantaged all level-50 stats yields wins on only a handful of additional early-game bosses and leaves the rest unwon. The failure cases range from sub-10-second deaths in cramped, multi-target encounters to minute-long stalemates that inflict almost no damage, and we report them through survival time and damage dealt rather than win rate alone.

Large language model evaluations typically focus on performance under nominal conditions, creating an illusion of capability where models comfortably walk a narrow, highly optimized generation corridor. In real-world deployments, however, complex system prompts, safety guardrails, and structural constraints continuously force models off this nominal path, driving a divergence between benchmark scores and deployment performance. To address this issue, we introduce Decoding-Level Taboo, a zero-prompt diagnostic stress test that intervenes directly in logit space at runtime, forcing models out of their nominal paths. By dynamically masking primary candidate tokens at word boundaries, Taboo forces machine circumlocution. Evaluating Taboo across several open-weight model families reveals that off-path robustness is heavily influenced by both parameter scale and post-training instruction alignment, with robustness generally improving with model size and alignment. Beyond the results presented in this paper, Taboo provides a novel primitive for generating diverse synthetic datasets, stress-testing runtime safety guardrails, and auditing model reliability prior to real-world deployment.

In fair ranked link prediction, demographic parity ($Δ_\mathrm{DP}$) is a common fairness metric. Yet, Mattos et al. (2025) argue that it fails to detect exposure bias because it ignores where links appear in the ranking. In this study, we reproduce this claim by showing that $Δ_\mathrm{DP}$ can indicate aggregate parity even when some subgroup-pair links are systematically ranked lower than others. The proposed rank-aware Normalized Discounted KL-divergence (NDKL), however, does detect such disparities. We also reproduce the effectiveness of MORAL, a post-processing method that improves exposure-based fairness while maintaining competitive utility. Beyond reproduction, we assess robustness using synthetic homophily settings, categorical sensitive attributes, and additional fairness and utility metrics, including subgroup-pair-adapted Attention-Weighted Rank Fairness (AWRF). Overall, our results show that exposure-based metrics uncover biases hidden by $Δ_\mathrm{DP}$ and that MORAL reduces these biases with minimal utility loss across diverse settings and datasets. We release a corrected, reproducible implementation at https://github.com/Floris93100/reproducing-MORAL.

Test-time scaling often uses an external verifier, such as compilers and test cases in coding or trained value functions in robotics applications, to obtain high-quality rollouts. Verifier-free test-time scaling (or VF-TTS) is gaining extensive attention as a mechanism to enhance Large Language Model (LLM) reasoning, primarily because we do not have access to such high-quality verifiers in many real-world applications. Among existing VF-TTS methods, confidence-based VF-TTS methods, which compute and rank rollouts solely by confidence, are particularly promising. Such methods introduce near-zero overhead for sample evaluation and require minimal access to internal model states, making the methods highly flexible across models and tasks. In this paper, we demonstrate a critical limitation of existing confidence-based VF-TTS methods by showing that such methods catastrophically break down on complex tasks. We observe a very interesting phenomenon: uniformly high confidence frequently indicates a failure to explore, favoring confidently wrong answers. To address this, our core insight is that robust cognitive search requires a specific confidence trajectory pattern: such methods perform exploratory branching at the beginning, as manifested by low initial confidence, and converge to a high final confidence solution. To implement this insight, we introduce consilience, a novel selection framework that explicitly evaluates the temporal asymmetry of confidence in reasoning. We operationalize this via a combinatorial metric that actively penalizes high initial confidence while strictly demanding final certainty. Extensive experiments covering both graduate-level mathematics problems and free-form code generation demonstrate that consilience effectively outperforms existing baselines, validating our novel perspective on completion confidence.

Thinking Mode Fusion (TMF) enables large language models to support both concise responses and long-form reasoning by unifying a non-thinking mode and a thinking mode within a single model. However, its training dynamics, including the \emph{data ratio} and \emph{training schedule} between the two modes, remain underexplored. In this work, we present a systematic study of TMF by analyzing the effects of the training schedule and data ratio between thinking and non-thinking modes. Focusing on mathematical problem solving, we construct a benchmark with multiple thinking-to-non-thinking data ratios and three training schedules. Our results reveal an asymmetric interaction between the two modes: increasing the ratio of non-thinking supervision reduces the accuracy of the thinking mode. We further show that different training schedules modulate this trade-off and that the optimal schedule depends on the data ratio. Finally, we quantify a negative correlation between non-thinking and thinking mode supervision, highlighting an inherent tension between these two modes. These findings provide practical guidance for designing effective TMF training settings. All code and data are released to support further research at: \href{https://github.com/caocongfeng/Fusion-Bench.git}{\textbf{Fusion Bench}}.

We introduce BDH-CQ, a reasoning model that combines in-context learning with recurrent latent reasoning. Inputs presented at inference time continuously update the model's recurrent memory; the model then solves a query through iterative computation in a high-dimensional latent space, without verbalizing its intermediate reasoning. We evaluate the model on the public ARC-AGI-1 evaluation set and use controlled ARC-like interventions to study what it learns from demonstrations, how consistently it applies an inferred transformation, and which concepts remain difficult. A 150M-parameter configuration reaches 29.5% pass@2 at a computed inference cost of \$0.0007 per task. This operating point breaks through the previously reported ARC-AGI-1 cost-accuracy Pareto frontier, establishing a new state of the art in benchmark cost efficiency.

Ball possession is the most-cited and most-misleading number in football: 60% recycled in one's own half is not 60% spent pinning the opponent back. Existing event-based possession-value frameworks (expected threat, VAEP, on-ball value) price on-ball actions but ignore the off-ball question a sterile possession poses: did holding the ball create space, or was the circulation dead? We answer this in two layers. First, an event-side junk-possession index prices each possession sequence by its peak threat gain under an expected-threat grid and -- after reconstructing the live scoreline to exclude lead-protecting circulation -- flags low-threat sequences in tied-or-losing states. On the 2026 FIFA World Cup (103 matches, 206 team-matches) the flag correlates negatively with points (r=-0.37) and xG difference (r=-0.51, partly index-coupled). It is not a repackaging of on-ball value: with team offensive VAEP and field tilt held fixed, the junk flag stays strongly negatively associated with points (p<0.0001, also match-clustered) while VAEP is not significant -- in this same-match (descriptive) regression it adds information beyond this on-ball action-value model. Second, for a flagged window we resolve whether it was spatially dead or space-creating by projecting broadcast video to pitch coordinates and measuring a Space-Creation Index (SCI): a net pitch-control change capturing whether the possession seized space or pushed the opponent's block back. Across 31 of 35 flagged windows from nine World Cup matches (a purposive sample), 74% are spatially non-space-creating, 19% weak progression, and 6% space-creating windows the event flag alone would score as failure -- including a side with 73% of the ball that exited on penalties (two non-creating windows). The two layers separate space-creating-but-unconverted from sterile possession, a distinction event-only on-ball value cannot make.

The safety of large language model (LLM) agents depends not only on model weights but also on the agent harness that manages context, memory, tools, permissions, and runtime control. Existing safety mechanisms often treat the harness as a fixed deployment artifact, limiting their ability to evolve with emerging risks. Moreover, coupled functions across harness components obscure safety responsibility attribution, making localized evolution difficult. We propose Safety Harness Evolution (SHE), a framework that learns evolving safe boundaries from rollout trajectories. SHE decomposes the harness into four artifacts with explicit safety responsibilities, including the System Prompt, Rule Bank, Safety Memory, and Tool Policy, defining clear functional boundaries for localized evolution. Based on this decomposition, SHE introduces an attribution-guided evolution loop that converts trajectory failures into structured diagnoses, learns artifact-specific boundary refinements, and selects evolved harnesses through safety-utility validation. Experiments on Agent-SafetyBench demonstrate that SHE effectively enhances safety through harness evolution, achieving a 3.1x ASR reduction compared with static SafeHarness, while also improving benign utility. The evolved harness further generalizes to unseen risks on the held-out AgentHarm benchmark and transfers across agent models without additional evolution.

Financial prediction typically relies on task-specific regression, ranking, or policy heads, separating the language model from the numerical object ultimately evaluated. We investigate whether a causal language model can instead represent forecasts and decisions directly through constrained token generation. FinATOM introduces a unified, head-free interface for three-step stock-return forecasting and dynamic five-ETF allocation. The forecasting model autoregressively emits volatility-standardized return tokens and is trained with ordinal and ranking supervision followed by a one-epoch token-level policy stage. The allocation model generates normalized long-only weights; supervised fine-tuning imitates a causal mean--variance anchor, and DAPO-augmented GRPO optimizes realized 21-day Sharpe subject to anchor consistency. In 2023--2025 ETF tests, the allocation policy improves pooled gross Sharpe from 1.428 to 1.529 and net Sharpe under a 5-bp transaction-cost model from 1.394 to 1.494. The multimodal allocation input attains the highest three-period mean Sharpe of 1.540, with its clearest advantage in 2025. On FinTexTS, the SFT and policy strategies achieve 73.52\%/2.68 and 73.72\%/2.69 cumulative-return/Sharpe, respectively. These results support the feasibility of direct language-model token generation for financial numerical prediction and decision-making, while motivating broader tests across assets, regimes, and random seeds.

Physically consistent motion planning remains a fundamental challenge in embodied AI, as generated trajectories must strictly conform to real-world execution dynamics. While latent world models offer a promising approach by predicting these dynamics, existing methods learn unconstrained future representations where absorbed physics remains implicit. Therefore, they fail to form reusable physical knowledge, which compromises reliability in unpredictable open-world navigation. To address this, we propose a novel Energy-Structured Latent World Model (ELWM). Our key idea is to structure the ELWM latent state to explicitly carry energy and momentum, ensuring strictly causal transitions via dissipation and control ports. Trained on multimodal RGB-D and inertial interaction histories, our model guarantees physically consistent predictions. We further implement this for motion planning by constructing Physics-Conditioned Neural Time Fields (PC-NTF), a key technical cornerstone that integrates ELWM into an arrival time field via the Eikonal equation to yield a physically-informed navigation policy. Across held-out scenes, our evaluation reveals significant improvements. Compared to generic latent models, PC-NTF reduces 0.8-s motion-prediction NRMSE from 0.36 to 0.29. Against Active Neural Time Fields, it improves navigation success from 81.3% to 89.7% and SPL from 0.64 to 0.73, while cutting the physical collision rate from 12.1% to 5.8% and the Eikonal residual from 0.083 to 0.031. Beyond these targeted gains, our results demonstrate that embedding explicit physical structures into latent spaces intrinsically bridges the gap between predictive world models and safe, dynamically feasible motion planning.

Code comprehension is central to software maintenance and evolution, yet different Python repetition structures may impose distinct cognitive demands. We conducted a controlled eye-tracking experiment with 32 undergraduate students with prior Python experience to compare comprehension of for loops, while loops, recursion, and list comprehensions (LCs). Participants solved six comprehension tasks in a Latin Square design while we measured completion behavior and eye-tracking metrics over full snippets and construct-specific Areas of Interest (AOIs). for loops showed the lowest visual effort. Relative to for, while loops increased AOI fixation duration by up to 97% and regression count by 114%, with regressions concentrated around counter management. Recursion showed a descriptive 50% increase in regressions, mainly between the base case and recursive call. LCs increased AOI time by 62.5% and fixation duration by 80.9%, with horizontal regressions suggesting dense token-by-token parsing. LC comparisons yielded the clearest statistically significant pairwise differences, while the combined comparison of all non-for structures was significant across all eye-tracking metrics. These findings provide process-level evidence that Python repetition structures induce distinct visual-effort patterns, with implications for readability, code review, refactoring, onboarding, and maintainability.

Agentic artificial intelligence has shown great promise in automating algorithm design, but scaling similar techniques to computer microarchitecture discovery remains challenging due to vast search spaces, strict hardware budgets, and long simulation times. In this work, we present ArchAgent v2, a framework which scales automated microarchitecture search to multi-level data prefetching. While the original ArchAgent successfully discovered single-level cache replacement policies in competition settings, it does not scale to multi-level prefetching where the design space and degrees of freedom are larger. To overcome this, we introduce two new additions to ArchAgent: a cascaded evolutionary search that subdivides the design space by sequentially evolving and freezing prefetchers at individual cache levels, and a hardware-realizability feedback loop that embeds real-time size-estimation directly into the evolution process. Evaluated under identical rules of the 4th Data Prefetching Championship (DPC4), ArchAgent v2 automatically designs a three-level prefetcher that outperforms the winning hand-designed solution, further demonstrating automated agentic discovery as a useful tool for computer architects. Our discovered policy achieves a 3.8\% geometric mean IPC speedup over the baseline overall and a 0.3\% improvement over the prior champion, BertiGO. On low-bandwidth single-core configurations, our policy yields a 4.6\% performance speedup compared to only 2.6\% for BertiGO. However, multi-core evolution still remains a significant challenge due to simulation latency impeding evolution speed. Finally, our profiling of an ArchAgent evolution of over 12,000 candidate designs provides key insights into how automated evolutionary agents explore and synthesize complex microarchitectural logic.

We introduce Sci-VBench, a comprehensive benchmark for evaluating knowledge- and reasoning-intensive video generation across scientific domains. It contains 1,253 expert-annotated examples spanning 60 subjects across four core disciplines: Natural Science, Healthcare, Humanities & Social Sciences, and Engineering. Each example requires models to generate temporally rich videos that demand scientific reasoning and knowledge-grounded synthesis, going beyond surface-level visual plausibility. We further establish a rubric-based evaluation protocol. Our analysis shows that, under this protocol, both non-expert human evaluators and MLLM-as-Judge systems can achieve relatively high agreement with expert judgments, supporting reproducible evaluation at scale. We benchmark 16 frontier proprietary and open-source models and find that, while automatic perceptual-quality scores cluster tightly across systems, performance on Prompt Grounding and Scientific and Causal Correctness varies substantially, with a pronounced proprietary-open-source gap. These findings show that advances in visual realism have not yet translated into reliable modeling of scientific and causal dynamics.

Uniform stability is a classical tool for controlling the generalization error of a learning algorithm. Bousquet, Klochkov, and Zhivotovskiy (2020) showed that the problem can be reduced to a moment inequality for a sum of weakly interacting functions of independent random variables. Their bound contains an additional factor $\log n$, and they asked whether this factor can be removed. We answer this upper-bound question affirmatively. More specifically, let $Z=(Z_1,\ldots,Z_n)$ have independent coordinates and let $g_i(Z)$ satisfy $$ \mathbb E[g_i(Z)\mid Z_{-i}]=0, \qquad \left| \mathbb E[g_i(Z)\mid Z_i]\right|\le M, \qquad \forall i = \overline{1, n} $$ while changing any coordinate $Z_j$, $j\neq i$, changes $g_i$ by at most $β$ and $Z_{-i}$ denotes all coordinates except $Z_i$. We prove that, for every $p\ge2$, $$ \left\| \sum_{i=1}^n g_i(Z)\right\|_p \le 16pnβ+M\sqrt{2pn}. $$ This removes the $\log n$ factor from the previous bound and matches the lower bound of Bousquet, Klochkov, and Zhivotovskiy up to universal constants in the range covered by their construction. Our proof first establishes the required estimate on the Rademacher cube, then transfers it to arbitrary product distributions by a two-copy randomization argument.

Leading large language model providers now conceal their models' step-by-step reasoning, or chain-of-thought, to protect intellectual property and limit information leakage. Rather than storing these traces server-side, providers return them to the client as blocks of encrypted text, which the client passes back with each subsequent request. Building on prior research, we identify an architectural vulnerability: these encrypted blocks are fully compatible and interchangeable across different sessions, users, and models within a provider's ecosystem. We exploit this compatibility to develop a scalable decryption jailbreak. By injecting an encrypted reasoning trace from a given model into a weaker, and less safeguarded model from the same provider, we force it to decode and output the trace verbatim in plaintext, without ever jailbreaking the more capable model directly. This vulnerability enables four distinct attack vectors. First, it circumvents anti-distillation mechanisms, allowing adversaries to extract a proprietary model's reasoning, as we demonstrate across Anthropic, OpenAI, and Google. Second, it allows for large-scale private data extraction. Developers frequently share session logs publicly, unaware of contents of the encrypted blocks. By decoding 315,320 reasoning blocks scraped from public repositories, we recovered 367 Personally Identifiable Information (PII) artifacts and 182 credentials. Third, it inadvertently reveals hazardous information hidden within the reasoning process, even in cases where the model's final, visible output safely rejects a malicious request. Fourth, attackers can leverage this flaw to execute invisible prompt injections, embedding malicious payloads entirely within encrypted blocks to poison public agentic rollouts. Following responsible disclosure, we propose concrete cryptographic and system-level mitigations to secure client-side reasoning.

Audio-visual interaction is the standard for patient-physician consultations, enabling natural communication and effective assessment of illness through non-verbal cues. While text-based AI has shown promise, it discards essential perceptual dimensions and limits patients who cannot articulate symptoms in writing. Early efforts to extend medical AI to audio-visual interaction have demonstrated feasibility but not reached clinician-level performance. Here, we provide the first demonstration of expert-level AI in real-time clinical video consultations using AMIE (Articulate Medical Intelligence Explorer) in a video configuration. AMIE (Video) is a Gemini-based multi-agent system integrating low-latency dialogue, clinical reasoning, and real-time audio-visual perception. To guide development, we established a taxonomy and automated evaluations for clinical audio-visual cues in telehealth settings. In a randomized Objective Structured Clinical Examination (OSCE) study with 30 primary care physicians (PCPs), 15 patient actors and 100 clinical scenarios, we compared AMIE (Video), its text-only counterpart AMIE (Text), and PCPs consulting via video. Clinical evaluators rated AMIE (Video) on par or better than PCPs in history-taking, diagnosis, management, and physical observation and examination. Patient actors preferred AMIE's approach to assessing and explaining conditions, while PCPs were preferred for rapport and partnership building. In modality ablation, patient actors preferred AMIE (Video)'s interface over text chat for communicative effectiveness, convenience, and feeling understood. Limitations remain in fine anatomical precision, subtle affective nuances, and high-frequency movements. While further research is needed before real-world translation, these results mark an important milestone toward AI systems capable of augmenting care across the sensory complexity of clinical practice.

Advances in advanced artificial intelligence tools have sparked research in robot autonomy, but the development of such systems has largely focused on execution rather than verifying the feasibility actions planning models propose. Like general-purpose LLMs, robotics planning models carry risks: biased toward user-specified goals, they may suggest actions misaligned with scientific ethics, they may be unsafe due to an inability to "remember" prior safety risks, or they may be vulnerable to adversarial attacks on the autonomy ecosystem. We propose a LLM-driven verification layer between planning and execution to evaluate action permissibility. Our LLM-as-a-Judge ensemble combines chain-of-thought reasoning across models and synthesizes those expert judge outputs, mirroring a combination of a mixture of experts and self-consistency approach. This layer serves as middleware, gating plans from the server's planning module before they reach the MCP server and therefore the robot's low-level controls: plans are approved, rejected for reformulation, or escalated for human review. With this system, we achieve near 85% precision across accept/escalate/reject categories 97% containment of adversarial attacks, with negligible errors between accepting and rejecting tasks, and errors mostly manifesting at the escalate boundary.

Autonomous research agents can generate experiments faster than researchers can validate them. Researchers have responded by scaling the proposer and ranking more samples with a learned judge or human reviewers. We argue that this *generate-and-rank* paradigm misses the problem of sparse feedback. Within a declared research problem, an agent follows the control loop of a greybox fuzzer: it proposes a candidate, executes it, observes feedback, and chooses what to try next. A fuzzer rarely finds a bug, but coverage makes partial progress observable on every execution. Fuzzers then use that signal to mutate inputs and allocate effort, rather than only to rank completed runs. Auto-research needs the same two capabilities. First, each experiment should expose a cheap, dense signal of epistemic progress before final scientific validation is available. Second, that signal should determine the next intervention so that the agent searches rather than repeatedly samples. Because the optimized progress signal is guidance rather than a verdict, final validation must still decide what counts as a discovery using evidence protected from adaptive reuse. We propose controlled tests of whether candidate signals predict validated progress, whether feedback-directed search yields more validated discoveries per unit cost than repeated sampling, and whether protected validation reduces false discoveries. Feedback architecture, not only generation, is a central bottleneck in auto-research.

General-purpose reward models are increasingly the bottleneck for scaling robot learning, yet the recipe for learning value-related capabilities from large-scale heterogeneous corpora remains underexplored. Existing approaches tie supervision to task-internal anchors such as preferences or normalized progress, none of which transfer cleanly across embodiments and data sources. We introduce RynnValue, an open-source value foundation model for robotic manipulation that replaces these anchors with temporal distance, the directed cost-to-go from an observation to the language-specified goal. Because temporal-distance labels can be derived directly from timestamps, RynnValue scales to over 7,000 hours and roughly 3M instruction-conditioned clips without preference or progress annotations. To make temporal-value learning reliable at scale, we combine random temporal sampling, temporal-order shuffling, and value-isolation attention, suppressing shortcuts that would leave predictions insensitive to failures and regressions. Trained without preference labels, RynnValue attains an average Kendall's tau_a of 0.675 on RBM-EVAL-OOD, surpassing the fully preference-supervised state of the art (0.655) and more than doubling a progress-only counterpart (0.292), while generalizing zero-shot to unseen tasks, embodiments, and viewpoints. Converted into dense rewards via potential-based shaping, it raises real-world policy success from 52.5% to 72.5% online and from 63.8% to 82.5% offline. These results establish temporal distance as a scalable supervision target and practical reward interface for generalist robot policies.

The development of embodied Intelligent Virtual Agents (IVAs) that have cognitive capabilities in real-time interactive virtual environments remains a challenge, even with today's advancements in technology. Existing architectures are often focused on either the implementation of low-level reactive control systems that are constrained by commercial game engines, or high-level representations of reasoning models that can be difficult to implement in virtual worlds. This paper builds on that notion and proposes a modular cognitive architecture for deploying embodied IVAs. This architecture builds on existing, pre-established frameworks such as the Sense-Think-Act paradigm and the Belief-Desire-Intention cognitive model, among others, and aims to provide a reusable implementation-oriented framework as a template for deploying IVA "brains" in interactive 3D computing systems. The proposed architecture contributes by providing a modular, implementation-oriented framework for the deployment of embodied, cognitive-capable IVAs and bridges the gap between high-level agent reasoning models with real-time embodied execution, for scalable, adaptive, and explainable agents in complex interactive virtual environments.

This paper presents a methodological framework for real-time climate risk assessment using data-driven nowcasting techniques to enhance supply chain resilience in Colombian agricultural contexts. Climate variability in Colombia, characterized by irregular rainfall, temperature fluctuations, and recurrent extreme events, has a direct impact on agricultural production and logistics, particularly for time sensitive crops. The proposed approach integrates short term climate forecasting based on historical meteorological observations with supply chain risk modeling to establish a conceptual early warning system architecture. A prototype implementation developed in a controlled computational environment demonstrates the feasibility of the framework using historical meteorological and agricultural time series derived from official statistics and reanalysis products, without reliance on satellite imagery or computer vision components. The methodology addresses the integration of climate nowcasting with supply chain decision making through explicit risk mapping, threshold-based categorization, and stakeholder-oriented risk signals. Results from synthetic and historical data experiments indicate that short term precipitation nowcasts can be translated into actionable risk indicators for agricultural supply chains, supporting anticipatory decisions related to inventory, sourcing, and transport.

On-policy distillation (OPD) has emerged as a core component of modern LLM post-training pipelines, yet we reveal a failure mode: degenerate agreement, where students exploit repetitive loops to achieve near-perfect token agreement with the teacher despite globally flawed responses. We therefore shift our focus from agreement to teacher-student mismatch, and find that mismatch tokens can be mainly categorized into two types: student-excess tokens and student-deficit tokens. Student-excess tokens are generated by the student but assigned near-zero probability by the teacher; their log-ratio corrections grow unbounded and destabilize the update. Student-deficit tokens, in contrast, are preferred by the teacher but rarely sampled by the student; their absence blocks the transfer of the teacher's reasoning patterns. To tackle these mismatch directions, we propose TIDE (Token-level Independent Deficit-Excess correction), which applies bounded Hellinger shaping to suppress the most severe sampled excesses and an analytic teacher top-$K$ injection to restore deficient probability mass without requiring deficit tokens to be sampled. Across mathematical reasoning benchmarks with multiple Qwen3 teacher-student pairs, TIDE consistently outperforms standard OPD and recent token-selection and reward-shaping baselines. Moreover, the gains of TIDE are more pronounced under strong teacher-student mismatch, where it improves Avg@8 from 6.9% to 20.3%, reduces average response length by a factor of 3.6, and substantially reduces formatting failures. Code is available at https://github.com/yzc-666/TIDE

Financial sentiment analysis converts unstructured financial news into quantitative signals that can support market analysis and decision-making. Existing work on resource-efficient financial NLP has largely focused on compressing or adapting pretrained language models, with less attention to combining contextual representations with lightweight rule-derived features. This study develops Rule-Aware FinBERT (RA-FinBERT), a parameter-efficient framework that integrates low-rank adaptation (LoRA) with three continuous VADER-derived sentiment proportions (positive, negative, and neutral) and a source-level metadata feature. The standardized four-dimensional feature vector is directly concatenated with the 768-dimensional final-layer FinBERT [CLS] representation and passed through a lightweight classification head. This design introduces only 1,024 additional trainable weights relative to a structurally matched text-only FinBERT model. RA-FinBERT was evaluated against text-only FinBERT and a lightweight DistilBERT baseline for three-class sentiment classification of financial-news titles and descriptions. On the held-out test set, RA-FinBERT achieved 69.89% accuracy and a macro F1 score of 0.634, compared with 63.44% and 0.526 for text-only FinBERT. Neutral-class recall increased from 18.18% to 45.45%. The framework supports both CPU and GPU execution, offering a lightweight and practical approach to financial sentiment classification under constrained computational resources. These findings indicate that rule-derived sentiment information and source metadata can provide complementary signals to contextual FinBERT representations and improve performance with minimal additional model complexity.

Body-focused repetitive behaviors, such as hair pulling and skin picking, are compulsive motor actions commonly associated with obsessive-compulsive and anxiety disorders. Their early, objective detection remains difficult because the movements are subtle and overlap with ordinary, non-pathological gestures. We developed and evaluated a multimodal deep learning framework to detect and classify these behaviors from wrist-worn sensor data. The data, collected by the Child Mind Institute using the Helios wrist-worn device, combine inertial measurement units, thermopile sensors, and time-of-flight sensors, capturing kinematic, thermal, and proximity information. The framework combined a convolutional neural network with a gated recurrent unit, alongside modality-specific autoencoders and a late-fusion classifier, to exploit temporal and spatial dynamics. It achieved an F1 score of 0.985 and an area under the receiver operating characteristic curve of 0.997 for binary detection, distinguishing these behaviors from other activities, and a macro-averaged F1 score of 0.700 with an area under the curve of 0.963 across a nine-class scheme that distinguished each individual behavior from a single grouped Non-Target class, improving over single-modality baselines. Post-hoc interpretability based on Shapley additive explanations showed that the time-of-flight and inertial modalities dominated discriminative power by capturing spatial proximity and dynamic movement, while hierarchical clustering indicated that misclassifications were driven primarily by the anatomical region of the gesture. These findings demonstrate that multimodal sensor fusion enables accurate, objective, and continuous behavioral monitoring. This work establishes a foundation for real-time, wearable-assisted mental health diagnostics and personalized interventions in biomedical research and clinical care.

AI agents increasingly work inside systems that govern how they delegate tasks, move information, execute actions, and use shared resources. Recent work already shows that deployment rules can change collective behavior. Here we ask which parts of an AI institution produce safety and how they do it. This is the first paper from POLIS, an ongoing research programme studying algorithmic institutions for multi-agent systems. We report a frozen 5,280-episode study suite. The main pre-specified delegation experiment spans four model families; a targeted high-conflict diagnostic adds three additional model endpoints. In matched structured workflows, the model sees different rule formulations and guards consult different authority states. We also vary the attractiveness of the immediate compliant internal/self fallback and allow blocked workflows to continue. A detailed constitutional prompt produces 0/384 realized violations. A provenance-aware executable guard also produces 0/384, although it blocks prohibited attempts in 51/384 episodes; 44/51 of those episodes later complete safely. The local-state guard's failures concentrate in scenarios where an ordinary transformation changes visible policy while originating authority stays fixed. In matched laundering scenarios, that guard admits violations in 22/96 episodes and provenance enforcement in 0/96 (p = 4.77 x 10^-7). A separate resource-allocation experiment shows that revealing the numerical value of an otherwise identical cap changes agent requests. In these structured workflows, the same final violation rate can hide very different mechanisms. The rule itself is only part of the institution. The authority state the system trusts matters, and so does the path available after a block.

Reinforcement learning with verifiable rewards yields no group-relative signal when rollout groups are uniformly correct or uniformly wrong, which account for 63.0-68.0% of groups in our experiments. We propose SKALD (Skill-Anchored Latent Distillation), an on-policy self-distillation framework that uses two context views of the same Qwen3-Base model: a question-only student and a teacher conditioned on an abstract, explicit-answer-filtered skill card. The student is trained on its own prefixes, transferring the skill-induced advantage into shared parameters without privileged input at test time. To stabilize context-induced distribution mismatch, SKALD employs an annealed exponentially tilted objective that downweights teacher-preferred tokens with very low student likelihood; as the tilt vanishes, it converges to teacher cross-entropy and recovers the forward-KL student gradient. An empirical gate activates distillation only when verified rollouts estimate a positive teacher advantage. Across five held-out mathematics benchmarks, SKALD improves overall avg@8 over GRPO by +2.46, +4.85, and +12.01 at 0.6B, 1.7B, and 4B, respectively. At 1.7B, zero-variance-only distillation recovers 84.7% of the full gain, while SKALD remains +4.06 above FLOP-matched GRPO and exceeds contextual skill exposure by +3.77. These results show that abstract skills provide dense supervision where group-relative rewards become uninformative.

Macaron-V1 is an open agent-model family for experiential intelligence: learning from experience in real environments and continuing to learn after deployment. It is organized around two system goals. Adaptation is pursued through recursive improvement of versioned model-harness pairs, where experience from one configuration is evaluated under an external contract and used to construct its successor. Collaboration is pursued via the Mixture-of-LoRA (MoL) architecture that freezes a base model, composes specialist LoRA adapters, and selects one LoRA per user turn. The flagship Macaron-V1-Venti combines a 744B GLM-5.2 base with four LoRAs for chat, agent, coding, and GenUI; the Qwen3.6-based Macaron-V1-Tall (50B) uses the same design for local deployment. This report presents Macaron-V1 as a co-designed system spanning architecture, algorithms, and infrastructure. The MoL architecture supports continual learning through extensible LoRA specialists. The algorithm combines Model-Harness Co-design and recursive self-improvement loop, including the UI4A component-native GenUI harness, a stateful action substrate, versioned HCP contract, and the agentic RL framework MindForge. The supporting infrastructure includes the post-training platform MinT, the long-context RL method LongStraw, and stability techniques for sparse MoE and DSA base models. We evaluate Macaron-V1 on Personal Intelligence, GenUI, and general capability benchmarks against frontier baselines. Our results validate the current system, while compounding gains from continual learning and collective intelligence remain open questions.

Reliable medical image understanding requires models to connect clinical language and visual reasoning with pixel-level grounding. Yet medical vision-language models often lack precise localization, whereas medical segmenters typically rely on explicit target categories or precise spatial prompts. This divide is reinforced by a supervision mismatch: segmentation datasets provide precise masks but little language supervision, whereas medical vision-language data rarely pair language with dense spatial annotations. To address this gap, we present MedPixel, a unified medical pixel-language model built around a shared language--mask interface. To provide scalable supervision, we introduce MedPLG-440K, comprising approximately 440K pixel-language task samples constructed through a clinically motivated synthesis process without external LLM annotation. MedPixel is trained with joint multi-task supervised fine-tuning followed by Pixel-Level Preference Optimization, which uses ground-truth masks as offline verifiers to derive response preferences from mask quality. MedPixel supports a broad spectrum of tasks spanning explicit grounding, implicit reasoning, spatial interaction, grounded explanation, and medical VQA. Across this task spectrum, MedPixel achieves strong performance in both pixel-level prediction and response generation, together with effective zero-shot transfer to external grounding benchmarks and robustness to imperfect spatial prompts. Code and model checkpoints will be released at https://github.com/yhy-whu/Medpixel.

Exploration has been a focus of reinforcement learning research for a long time. Recently, there has been growing evidence that it is also an important ingredient in LLM reinforcement learning recipes that can significantly impact downstream performance. Many existing methods control exploration in the action-space, for example, using temperature scaling. However, these methods cannot reorder tokens but only influence the variance in the output distribution. This limits exploration and can lead to divergence or stalled training. Here, we investigate parameter-space exploration, where rollouts are generated by sampling different policies from a posterior that may each explore different rollouts. Sampling less or more diverse policies is then a complementary control lever over exploration. We introduce a family of methods called Perturbed Parameter Policy Optimization (3PO) which use different sampling strategies and different rollout grouping for reward estimation. Experiments on OLMo-3-1025-7B and Qwen2.5-Math-7B across mathematical reasoning and code generation tasks show that these approaches consistently improve average downstream performance over standard GRPO at a near-identical FLOPs cost. Moreover, using multiple parameter samples consistently produces fewer zero-advantage groups and malformed or incorrect rollouts during training than GRPO and action-space baselines. Overall, our work presents evidence that parameter-space exploration can improve reinforcement learning for LLMs.

As AI coding agents take on increasingly complex, long-horizon software engineering tasks, existing benchmarks are rapidly saturating and their evaluation quality has come under serious scrutiny: a recent audit found that nearly 60% of unsolved SWE-bench Verified instances contain flawed tests -- either overly narrow tests that reject correct solutions or overly broad tests that check unstated requirements -- and that frontier models can verbatim reproduce gold patches from training data. Code refactoring, which requires coordinated, behavior-preserving changes across many files, offers a substantially harder and more realistic test of agent capability, yet remains underserved by current benchmarks. We introduce SWE-Bench ProMax, an expert-curated, multilingual code refactoring benchmark of 170 instances drawn from real commits across seven programming languages (Python, Java, TypeScript, Go, C, C++, and Rust). Every instance undergoes rigorous, multi-stage curation that directly addresses the quality problems identified in prior benchmarks: issue descriptions are rewritten from scratch to provide precise, unambiguous specifications, and test suites are manually reviewed to remove overly narrow and overly broad tests. Tasks with insufficient complexity or limited cross-file scope are filtered out, yielding a benchmark of challenging, large-scale refactoring tasks that average 11.4 modified files and 261.6 lines of code per instance, substantially exceeding the scale of existing benchmarks. Experiments with frontier models under two agent scaffolds show that the best model achieves only 41.2% resolve rate, confirming that SWE-Bench ProMax presents a meaningful and unsaturated challenge for current AI coding agents. Our benchmark is available at https://huggingface.co/datasets/swe-bench-promax/SWE-Bench-ProMax.

Joint exam-level prediction and candidate-region localization may improve the usefulness of AI support in mammography. We study this setting using a multi-task DETR framework, where shared representations support both image-level malignancy prediction and lesion localization, and evaluate its performance on OPTIMAM and a biopsy-confirmed SGM1k cohort. Across both datasets, modern backbones consistently outperformed older ResNet-style features, with ConvNeXtV2 and DINOv3 giving the strongest overall results, whereas MambaVision was less competitive. On OPTIMAM, ConvNeXtV2 achieved the best overall performance, reaching 97.96% AUC, 99.89% sensitivity, 25.08% mAP@.5, and 74.38% recall@.25. On SGM1k, DINOv3 gave the strongest overall results, with 90.97% AUC, 86.28% sensitivity, 82.00% specificity, 27.04% mAP@.5, and 77.32% recall@.25. These findings suggest that backbone quality is a critical factor in effective multi-task mammography, with ConvNeXtV2 emerging as a particularly strong and well-matched CNN backbone for mammography in this framework.

Modern coding agents increasingly appear capable of following complex software requirements, yet their success leaves a critical ambiguity: do they resolve the active specification, or merely follow the most salient path by which it was stated? We identify specification-path sensitivity, a failure mode in which requirement histories that are equivalent in their final meaning lead the same agent system to produce behaviorally different programs. This reframes evolving-requirement evaluation as active-contract resolution: before writing code, an agent must determine which requirements still count. Building on this view, we introduce SpecPath, a diagnostic evaluation that holds the repository, final contract, verifier, agent system, and execution budget fixed while changing only the revision path that leads to the contract. Rather than treating each patch as an isolated pass or failure, SpecPath uses paired executable outcomes to reveal whether an agent realizes the same tested behavior across contract-equivalent histories. Across five calibrated software tasks and fourteen coding-agent configurations, aggregate direct and revision-history accuracy is nearly unchanged; nevertheless, 35 of 100 complete blocks that succeed on the direct specification fail on at least one equivalent history. These results show that implementation success on a consolidated request does not guarantee specification-path invariance. Evaluating evolving requirements therefore calls for controlled tests of whether agents are robust to the path by which a specification becomes final.

Narrating the visual component of movies is known as audio description. It is a narrative technique designed to enable blind and visually impaired individuals to follow the story. However, it is far more constrained than most narratives: the descriptions not only need to convey the story in the movie, but they must also fit into gaps between dialogue and they need to conform to guidelines that exist in each region. In this work, we compare audio description created in the United Kingdom against audio description created in the United States. We use guidelines written for these two regions, alongside the impressions from a practitioner in the field, to motivate specific hypotheses about the differences. We test these hypotheses against our pre-existing corpus, which provides both human-authored American and British audio description for each of 206 movies. Results provide quantitative evidence to uphold all tested hypotheses, including differences in lexicon, the use of the progressive aspect, the use of passive constructions, the use of subjective adjectives and modifiers, when characters are named, how scenes are cued, and degree of overlap with movie dialogue and music. Our work offers a quantitative lens into the narrative technique of audio description.

Online credit card discussions provide a natural setting for studying how consumers communicate about financial products. Simulating these discussions requires more than just generating individual comments, the generated threads should also match how real users express themselves and interact with others. We introduce CARD, a framework for generating realistic credit card discussion threads. Given a credit card post and its matched real thread, CARD uses non-verbatim guidance on reply structure, comment function, stance, tone, and conversational variation. A planner organizes these controls, a writer generates the discussion, and a calibration loop updates comments' populations that contribute to differences between the generated and real thread distributions. We evaluate CARD on real Reddit credit card discussions using lexical, semantic, behavioral, and structural metrics. CARD matches the distributions of real credit card discussions better than simulation baselines across multiple LLMs and also demonstrates smaller effect sizes and distribution distances across metrics. These results show that structured planning and targeted revision can generate the realism of simulated credit card discussions.

Answering complex conditional questions using Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG) remains a challenge, particularly in domain-specific contexts where general-purpose LLMs and RAG tend to underperform. We hypothesize that augmenting RAG with unstructured and structured knowledge, extracted from both documents and knowledge graphs (KGs), can improve reasoning and answer accuracy for such tasks. To test this, we propose KGCaRe, a hybrid approach that combines neural retrieval with symbolic reasoning over LLM-generated KGs. KGCaRe constructs a KG from documents using a multi-prompt extraction strategy and stores it in a graph database. Simultaneously, the documents are embedded into a vector store to enable neural retrieval. KGCaRe performs innovative iterative graph traversal guided by the LLM to extract relevant triples, prune irrelevant information, and uses additional clue entities to traverse the graph again if the initial traversal does not provide satisfactory context to generate the answer. The relevant triples extracted from the KG in path form, along with semantically retrieved text passages, are then fed into custom KGCaRe prompts to generate answers to the complex conditional questions with explanations. We evaluate KGCaRe on two complex conditional QA datasets. Our results on these datasets show that KGCaRe consistently outperforms existing baselines, including Vanilla LLM, Code Prompt, Text Prompt, Think-on-Graph, Vanilla RAG, and HybridContextQA, across multiple LLMs such as Mistral, Mixtral, GPT-3.5, and GPT-4o. We publicly release the software pipeline that we developed to implement the proposed KGCaRe approach.

Accurate air quality forecasting is essential for public health and urban environmental management, but remains challenging because pollutant channels differ in periodicity and distribution drift, while their concentration trajectories contain both multi-scale dependencies and rapid changes. Recent methods have improved spatial dependency learning and meteorological covariate modeling. However, pollutant channels are still passed through the same normalization rule and temporal backbone, using a shared latent representation for channel-specific distributions and changes at different rates. To address this limitation, we propose AirFlow, a pollutant-aware dual-stream framework that operates on station multivariate observations without additional graph propagation or predefined signal decomposition. Specifically, AirFlow designs two novel blocks: (1) a statistic-guided normalization routing mechanism that selects a normalization path for each pollutant according to its 24-hour autocorrelation and distribution drift; and (2) a hierarchical dual-stream state model that combines multi-scale state space propagation with learnable response coefficients, where gated bidirectional cross-attention exchanges information and adaptively fuses the resulting representations. Experiments on real-world data from multiple cities show that AirFlow achieves the best performance in 34 of 36 metrics comparisons, with reductions of up to 11.11% root mean square error over the state-of-the-art baseline. AirFlow also requires only 0.0483M parameters and 0.0215G FLOPs, achieving high forecasting accuracy with low computational overhead.

Thoracic pathologies rarely occur in isolation, yet standard multi-label classifiers rely on shared global descriptors, discarding \emph{where} findings lie and \emph{how} they co-occur. We propose \textbf{C$\mathbf{^2}$A} (Co-occurrence Aware Class Attention), a classification head that explicitly couples spatial evidence with clinical priors. First, C$^2$A casts pooling as an expectation over learned per-class spatial attention maps, yielding localized descriptors for each disease. Second, it couples these descriptors via a learnable graph warm-started from empirical label co-occurrence. A single residual message-passing step shares evidence among related findings, proving to be a bounded perturbation of the identity where co-occurrence enters each logit through an explicit bilinear interaction. On CheXpert, C$^2$A achieves a superior $0.895$ macro-mean AUROC, outperforming advanced context-gating baselines. Crucially, gains concentrate on highly co-occurrent classes with ambiguous spatial evidence (rescuing Atelectasis by $+1.5$ over GCG), demonstrating the prior's regularizing effect with a negligible overhead of one linear projection and a $C\!\times\!C$ edge matrix.

Large Vision-Language Models (LVLMs) have demonstrated strong performance on multimodal benchmarks, yet it remains unclear whether they genuinely reason about relationships between images and text or rely on superficial correlations, known as shortcut learning. This question is particularly important for multimodal sarcasm detection, where successful prediction depends on recognizing pragmatic incongruity rather than treating sarcasm as simple image-text mismatch. We introduce PragMatch, a controlled benchmark of 3,000 image-text pairs derived from MMSD2.0, including original sarcastic examples and constructed literal and hard-negative pairs. We identify influential shortcut cues through systematic masking and evaluate their impact through targeted injection experiments. Our results show that LVLM predictions are sensitive to lexical, OCR-derived and stylistic cues, with injected surface signals causing substantial changes in model predictions despite unchanged underlying image-text relationships. Our findings reveal limitations in current LVLMs while PragMatch provides a systematic testbed for evaluating multimodal pragmatic reasoning beyond surface-level image-text alignment.

A prediction that is both confident and wrong is a critical reliability failure because it can bypass abstention and human review precisely when the model is mistaken. Empirical risk minimization (ERM) controls average loss but not this failure directly, while calibration, uncertainty estimation, conformal risk control, and selective prediction methods target related reliability properties rather than bounding the joint failure event during training. We propose ReliableNet, which constrains the Joint Confident-Wrong (JCW) probability, the probability that a prediction is simultaneously confident and incorrect, below a user-specified risk budget $α\in(0,1)$. We formulate this as a chance-constrained ERM problem, use a conservative smooth inner approximation whose population feasibility implies the original JCW constraint. Across four tabular and two image datasets, ReliableNet is the only method certified within the JCW budget for every dataset and seed in distribution, when compared against baselines spanning ERM, post-hoc calibration, conformal risk control, and selective prediction. Under demographic, ambiguity, spurious-correlation, novel-class, and covariate shifts, it achieves the lowest empirical JCW among the compared methods while remaining very competitive in accuracy, coverage, calibration, and selective prediction. Risk-coverage results further indicate that ReliableNet achieves better selective ranking than the benchmark methods on most datasets. Overall, ReliableNet provides a principled approach to trustworthy classification.

Real-time MRI makes it possible to observe vocal-tract articulation during speech, but mapping these articulatory patterns to phonetic and phonological categories remains challenging. We investigate whether PhonoQ, an audio-based model trained to recognize structured phonological features, provides useful information for audio--articulatory modeling. Specifically, we extract representations from PhonoQ's Conformer module, whose training is shaped by supervision for manner, place, voicing, and vowel features. Using articulatory contours with synchronized audio-derived features, we compare WavLM-large and HuBERT-large baselines with models that incorporate PhonoQ-derived representations. Across unseen-speech and unseen-subject settings, these features improve macro-F1 for phonological targets including manner, place, voicing, vowel height, and vowel backness, and also improve fine-grained 39-phoneme classification. In a contour-only inference setting, audio-derived teacher supervision yields modest but consistent gains over contour-only training, indicating that phonological information from synchronized audio can be partially transferred to articulatory models. Finally, posterior analyses show interpretable surface-sensitive patterns consistent with flapping-like /t/ realizations, /t/-/r/ retraction or affrication, and nasal place assimilation.

Multilingual translation benchmarks are typically sourced in English and translated into other languages, treating language pairs as the unit of evaluation---a design that is prone to contamination over time and overlooks locale and cultural considerations. We therefore advocate for source-contrastive evaluation and instantiate it with Cultivar, a localised subset of FLORES, which enables locale-specific translation evaluation. When paired with unlocalised counterparts, performance discrepancy allows the probing of data contamination and localisation robustness. We benchmark 32 open-weight models and find that MT-specialised models are less robust, a few models potentially overfit FLORES, and models tend to translate US content better than that of other locales, regardless of language.

Audio Description (AD) is a verbal narration of key visual content in videos, enabling access for visually impaired audiences. Unlike standard video captioning, AD is a structured editorial task: descriptions must be inserted into gaps in dialogue and must convey only what is needed to understand the narrative being told. However, existing approaches formulate AD generation in an artificial setting where both the content and timing of descriptions are pre-specified, reducing the task to clip-level captioning. They further rely on noisy transcription and alignment pipelines, and lack the rich parallel data required for modeling narrative context. We introduce a new formulation of AD generation in which models must jointly decide what to describe and when to do it. To support this, we present REFRAMED, a high-quality dataset of 2,023 videos that span 3,302 scenes from 206 movies, with professional AD transcripts (both American and British versions), professional subtitles and aligned screenplays. We also provide a manually curated challenge set that pairs full movies with multiple AD references, together with evaluation protocols that leverage dialogue gaps and multi-reference comparisons. Experiments with state-of-the-art AD systems and multimodal LLMs show that they outperform trivial baselines but fall far short of expert human performance. Our dataset and benchmark establish a new foundation for research on video understanding.

Transformer-based neural operators have achieved substantial progress in solving Partial Differential Equations (PDEs) by projecting spatial observations into compact latent tokens and learning physical interactions in latent spaces. However, we reveal that existing learnable projection mechanisms cannot ensure stable and balanced assignments from observation points to latent tokens, causing some latent tokens to be over-assigned while others remain underutilized. This limitation further restricts the design of hierarchical architectures, as assignment imbalance is continuously inherited and amplified across latent spaces, eventually causing severe token collapse in deeper spaces. To address these issues, we propose MoNo (Multiscale Optimal Transport Neural Operator), a progressive multiscale neural operator that efficiently solves PDEs on general geometries through stable latent-space construction. At its core is CoTAP (Cross-scale Optimal Transport Assignment and Projection), a novel latent-space construction method that formulates cross-space assignment between adjacent spaces as an entropy-regularized optimal transport problem, thereby constructing balanced bidirectional projections and stable latent spaces. CoTAP also ensures stable information transfer across multiple latent spaces, further enabling multiscale architectures on general geometries, which in turn support more efficient learning of long-range physical interactions. Extensive experiments demonstrate that MoNo outperforms existing state-of-the-art neural operators in both prediction performance and computational efficiency. Code is available at https://github.com/ZijiangY1116/MoNo.

Muon's polar update is exact for an unweighted spectral geometry. We introduce GO-MUON, which uses a matched data-dependent geometry and reuses it across several optimization steps. Conditioned on any positive-definite left and right maps, its raw update exactly solves the corresponding weighted spectral oracle; this statement is independent of how the maps are estimated or how recently they were refreshed. For softmax cross-entropy, we quantify when the observed-label backward factor approaches the model Fisher and generalized Gauss--Newton factor. We also show that four-step refresh nearly preserves the tracking delay of slowly changing geometry while increasing stationary factor noise, making lazy geometry a compute--statistics tradeoff rather than a denoising mechanism.

Retinal fundus images frequently exhibit multiple co-occurring pathologies, yet standard deep learning classifiers apply static, identical computation to every image regardless of the underlying disease distribution. We propose a novel architecture that resolves this via sparse conditional computation, pairing a Guided Context Gating (GCG) spatial attention front-end with a sparsely-routed Mixture-of-Experts (MoE) block operating over feature tokens. Crucially, this routing yields an interpretable, data-driven decomposition. Expert allocation is significantly disease-dependent (p < 0.001), with the healthy Normal state and morphologically distinct pathologies (e.g., ERM, AMD) isolating to dedicated experts. On a five-class, patient-disjoint 5-fold cross-validation benchmark, our model achieves 0.912 +/- 0.008 macro AUC and 0.653 +/- 0.014 macro F1. Furthermore, Grad-CAM++ and post-MoE t-SNE visualizations confirm that expert routing aligns with localized lesions and geometrically maps co-occurring cases between their constituent clusters, positioning sparse MoE as an interpretable approach to multi-disease retinal screening.

Decentralization as a concept in computer science has existed for over half a century. Despite its fundamental role across domains such as security, distributed computing, artificial intelligence, cloud infrastructures, and Internet of Things (IoT) architectures, there remains no universally accepted definition of decentralization applicable across computer communication systems. This has become increasingly problematic with the emergence of decentralized AI and machine learning paradigms, including collaborative training, distributed inference, blockchain-based, and agentic AI, where decentralization is often treated as a core design objective. Meanwhile, existing approaches frequently conflate decentralization with related notions such as distribution of trust or specific implementation paradigms. Such ambiguity creates inconsistencies in system analysis, limits comparability between works, and weakens the rigor of formal reasoning surrounding communication architectures and protocol design. In this work, we define this research gap as the Decentralization Problem. We analyze the formal-semantic, epistemological, and pragmatic foundations of decentralization and introduce a graph-based ontology defining it as both relational and subject-specific property of computer communication systems. The framework formally distinguishes decentralization from distribution and supports evaluation through two novel metrics: Void Tolerance and Imperviousness. We also provide a browser-based implementation that enables automated classification and metric computation of arbitrary systems. Instantiations to federated learning and blockchain architectures show consistent, comparable assessments where existing definitions produce incomplete or contradictory conclusions, providing a domain-independent foundation for analysing decentralization across heterogeneous systems.

On-policy self-distillation (OPSD) converts feedback into dense token-level supervision on trajectories generated by the policy to be optimized, providing a useful complement to reinforcement learning with sparse outcome rewards. However, the self-teacher policy used in OPSD is typically a stop-gradient or exponential-moving-average copy of the policy conditioned on additional context information, and thus co-evolves with both the student policy and its on-policy context distribution. Directly matching such a moving target with a fixed projection objective can lead to unstable optimization or excessive distributional concentration. This nature of OPSD motivates the proposed \emph{Self-Referenced On-Policy Self-Distillation (SR-OPSD)}. At fixed student-generated contexts, a token-level variational characterization identifies the effective distillation target as a geometric interpolation between the self-teacher policy and a reference policy. Meanwhile, we use the Rényi divergence family to generalize the projection geometry. This formulation separates \emph{where} the adaptive target is placed from \emph{how} the student is projected toward it: the interpolation coefficient controls underlying target, while the Rényi order controls the projection geometry and its sensitivity to token-level density ratios. Extensive experiments across scientific evaluation, mathematical reasoning, and coding generation tasks with multiple large language models show that SR-OPSD achieves the state-of-the-art or competitive performance across various settings.

Low-rank adaptation (LoRA) represents large language model (LLM) updates with two compact matrix factors, i.e., $A$ and $B$, providing an efficient way to fine-tune large models in federated learning paradigm. Inspired by the asymmetric roles of the LoRA factors, we study whether $A$ should be shared across clients while $B$ remains client-specific (Share-A/Local-B), or whether $B$ should instead be shared while $A$ remains client-specific (Share-B/Local-A). With a least-squares surrogate, we reveal that Share-A/Local-B requires the client-specific LoRA update matrices to use a common rank-$r$ input-side space, whereas Share-B/Local-A requires a common rank-$r$ output-side space. The two strategies therefore incur different projection residuals, indicating that the preferred strategy is the one with the smaller aggregate residual across clients. With this insight, we propose Federated Adaptive Factor Sharing Low-Rank Adaptation (FedAS-LoRA), which selects the sharing side before training to enhance fine-tuning performance. To enable adaptive factor selection before training, we design a Rank-Aware Shared-Subspace Sufficiency (RSS) metric, which effectively assesses whether a shared rank-$r$ input subspace is sufficient for the local data distributions using representations extracted from a frozen LLM backbone. Experiments across different tasks, data distributions, LoRA ranks, and participation settings confirm the effectiveness of RSS and the superior performance of FedAS-LoRA.

Large language models (LLMs) can generate functionally useful code that remains vulnerable, while security-focused interventions may break intended behavior. We investigate security tests as executable specifications both before generation and during iterative repair. We develop SecTDD, a controlled test-feedback scaffold that separates three factors: whether tests are shown upfront, whether failed executions trigger revision, and how failures are selected and represented. The evaluation uses behavior-partitioned visible and hidden tests and byte-identical initial candidates for repair comparisons. Across 2,705 trajectories, 31 task instances, three secure-code benchmarks, 16 CWE categories, and two model families, showing all visible tests upfront increases hidden functional-and-security joint success by 19.3 percentage points on average, but improves only seven of nine benchmark-model conditions and harms two. In shared-candidate comparisons, structured feedback repairs 80 initially unsuccessful candidates with no joint regressions; fixed raw feedback repairs 83 but causes three regressions. Structured and raw feedback are otherwise nearly indistinguishable head-to-head (six wins, six losses, and 453 ties). Candidates that pass all visible tests still fail hidden behavior families under every common regime. These results show that executable feedback can repair secure-code generation, but its benefits depend on the model, task, feedback entry point, and especially test coverage.

Agent skills are emerging as an important attack surface in LLM-based agent systems. Through an empirical study of existing skill scanners, we find that current defenses mainly inspect individual skills, leaving risks from cross-skill composition insufficiently examined. This creates a practical blind spot: multiple locally plausible skills may pass security checks while collectively forming a harmful workflow during agent execution. To investigate this threat, we propose ColluSkill, a collusive multi-skill-chain attack framework that decomposes a complete malicious intent into interdependent sub-payloads embedded in independently packaged skills. The attack does not rely on any single malicious skill, but emerges from the ordered composition of locally plausible behaviors through contextual dependencies, artifact passing, and execution handoffs. ColluSkill further employs LLM-based chain planning and scanner-feedback refinement to preserve chain-level attack semantics while reducing suspicious signals in individual sub-skills. To defend against such attacks, we propose ChainGuard, a context-aware skill-chain scanner that jointly analyzes a candidate skill and the skills already installed in the agent environment. ChainGuard reconstructs cross-skill dependencies, artifact flows, capability compositions, and downstream behaviors to identify risks that emerge only at the workflow level. Experiments on six representative skill scanners show that ColluSkill achieves an average attack success rate of 96.0% and consistently outperforms the evaluated single-skill and multi-skill attack baselines. Meanwhile, ChainGuard reduces the attack success rate to 22.5% while allowing 99.5% of benign workflows to pass, highlighting the importance of chain-level security analysis for agent skill ecosystems.

Lung cancer remains one of the leading causes of cancer- related mortality worldwide. Although targeted therapies have improved outcomes for patients with non-small cell lung cancer (NSCLC), they rely on mutation profiling through tissue biopsy, an invasive procedure with several limitations. This study investigates PET/CT-based radio- genomic prediction of epidermal growth factor receptor (EGFR), tumour protein 53 (TP53), and Kirsten rat sarcoma viral oncogene (KRAS) mutations using deep learning. We further evaluate whether pairwise multi-label learning improves mutation prediction compared with conventional single-gene classification. To the best of our knowledge, this is among the first studies to systematically investigate multi-label learning for PET/CT radiogenomic mutation prediction in NSCLC. Experiments were conducted on a novel UK-based radiogenomics cohort. Joint pre- diction of KRAS and TP53 improved AUC from 0.58 to 0.64 for KRAS and from 0.69 to 0.71 for TP53. For the EGFR/KRAS pair, only EGFR benefited from joint learning, while no improvement was observed for the EGFR/TP53 pair. These findings demonstrate that the effectiveness of multi-label learning depends on the specific combination of gene mutations being modelled, suggesting that mutation-specific modelling strategies may be preferable for PET/CT radiogenomic prediction.

Large language models (LLMs) are increasingly used to perform subjective evaluations traditionally made by humans, yet their validity as social judges remains unclear. This paper examines whether LLMs can assess social attraction from theory-grounded persona profiles constructed from ten psychological and relational constructs and organized into three tiers: socially attractive, socially mixed, and socially unattractive. We examine LLM ratings in two studies and compare them with human judgments in a third study. In Study 1, 34 LLMs rated 12 profiles across three repeated runs. Although some models tended to give higher or lower ratings overall, they showed strong stability across runs, consistent three-tier ordering, and high agreement in relative profile ordering. Study 2 examined sensitivity to gender presentation using six matched name-and-pronoun profile pairs and a separate pronoun-only test with a gender-neutral name, finding no significant effects in either analysis. In Study 3, 198 human participants evaluated the six matched profiles from Study 2. Their ratings reproduced the three-tier structure and followed a profile ordering consistent with that of the LLMs. However, LLMs rated attractive profiles more positively and unattractive profiles more negatively than humans, while neither group showed a significant overall effect of gender presentation.

Embedding trained neural networks as surrogates within optimisation problems is an established practice in operations research. The prevailing approach uses feedforward neural networks (FNNs) with ReLU activations, whose piecewise-linear structure admits an exact but computationally intensive mixed-integer programming (MIP) reformulation as the networks grow. We advocate input convex neural networks (ICNNs) as structurally superior surrogates when the underlying response is approximately convex or concave. The convex architecture offers two computational advantages. First, the ICNN-MIP formulation tends to yield a tighter linear programming (LP) relaxation than its FNN-MIP counterpart, with no integrality gap in favourable instances. Second, ICNNs uniquely admit an LP-based reformulation via epigraph representations of ReLU activations, though this embedding is not always exact. When it is not, we exploit the properties of ICNNs to construct the strongest continuous relaxation over box domains, namely, the convex hull of the ICNN's graph, bounded below by the epigraph and above by the concave envelope; this construction is tractable under input convexity but hard for general ReLU networks. On this basis, we develop a branch-and-bound algorithm that builds this relaxation at each node, branches directly on input variables rather than intermediate variables as in MIP reformulations, and terminates at the root node whenever the epigraph embedding is valid. Case studies on humanitarian food aid, oil well routing, and wine blending show that ICNN surrogates match FNN accuracy and deliver gains in solve time and scalability, supporting ICNN as the default surrogate when the underlying function is convex, concave, or well-approximated as such.

Large language models can write parametric CAD programs from a natural-language description (text-to-CAD generation), but a single sample is often wrong. Increasing test-time compute by sampling multiple candidates only helps if a good candidate can be identified, yet no ground-truth model is available at generation time. Existing systems often require a separate verifier, such as a vision-language judge, to select among candidates. We investigate whether the candidate pool itself provides enough signal for effective selection and a verifier-free alternative. We introduce 3D CAD consensus selection, hereafter consensus selection: sample $N$ parametric CAD programs, compile them to 3D models, and return the candidate that agrees most with the rest of the pool. The method is training-free and compatible with existing CAD agents. We investigate geometric and topological notions of agreement, each of which improves its corresponding evaluation metric. On the exact candidate pools of a state-of-the-art CAD generation method, geometric consensus improves all three geometric metrics over the method's verifier, while topological consensus matches it on topology. Across every tested LLM and prompt variant, geometric consensus also improves geometric accuracy over random selection from the same pool, reducing Chamfer distance by $1-10\%$.

Training a language model suite classically requires training each model separately and serving them independently. We improve both training and inference efficiency by stacking sub-models of increasing size into a single nested architecture trained end-to-end. This Matryoshka training framework reduces the total parameter count of the suite, enables low-cost distillation from the largest to all smaller sub-models at every training step, and is well-suited for speculative decoding as the draft model is contained within the verifier. We validate our approach by training a Matryoshka suite comprising 500M, 1.5B, and 3B sub-models. Our suite is on par with independently trained baselines on benchmark performance and validation and out-of-domain perplexities, while using 36% less training compute and improving the throughput of speculative decoding by 14-26%. We also ablate key architectural choices, offering guidance for building strong Matryoshka LM suites.

Great fiction earns its verisimilitude through precise details, from how a longsword is gripped to pierce armor gaps to why a bleeding corpse cannot yet smell of decay, weaving domain expertise into the fabric of invented worlds. Current AI writing tools offer limited support for discovering and integrating unfamiliar domain knowledge into narrative. They require explicit queries that authors cannot formulate, generate finished prose that risks homogenizing voice, or assist only within the boundaries of what authors already know. We argue that AI should reveal latent knowledge gaps to writers while preserving their agency to transform discovered knowledge into authentic prose. Grounded in formative interviews with 9 fiction writers, we present VeriForge, a mixed-initiative writing system that divides cognitive labor so that the system assumes initiative over domain discovery while the author retains full initiative over narrative synthesis. VeriForge realizes this through three complementary mechanisms. Proactive inline highlighting flags potential knowledge gaps as authors draft. Dual-stream querying pairs conversational responses with source-anchored Knowledge Cards for direct fact extraction. A spatial Knowledge Canvas allows authors to organize and connect discovered knowledge across their writing. These mechanisms are powered by a graph-based retrieval-augmented generation pipeline grounded in domain-specific source materials. A within-subjects user study (N=12) provides preliminary evidence that this paradigm helps authors recognize previously overlooked knowledge gaps, supports creative exploration, and is perceived by expert raters to produce passages with stronger domain grounding in a controlled cold-start writing task.

Predicting the answer to interventional ``what if'' questions --- the outcome of an action never taken --- requires a \emph{mechanistic}, causal model, not a curve fit; and learning such a model requires \emph{experiments}, because passive data leaves its mechanisms unidentified. Experiments are expensive, so the central problem is \emph{data efficiency}. We present the Model Discovery Agent (MDA), which couples a large language model (LLM), used as a \emph{proposer} of candidate structures, with standard Bayesian machinery --- sequential Monte Carlo (SMC) for parameter and structure posteriors, simulation-based inference (SBI) for intractable likelihoods, and value-of-information (VoI) for experiment design --- to discover latent mechanistic world models from few interventions. MDA operates in the M-open setting: when the truth lies outside the current hypothesis class, a predictive check flags the inadequacy and the proposer expands the hypothesis space with a new model whose parameters are then identified by designed experiments. We show that \emph{discovery and design reinforce}: the design step identifies the mechanism the discovery step proposes, and the identified mechanism improves predictions, enabling further discoveries from the remaining unexplained residuals. On three different benchmarks --- covering physics (\DPbench, \citep{wiemann2026discoverphysics}), chemistry (\CHEMbench, \citep{kabra2026autoscilab}) and biology (\HHbench, a new partially observed single-neuron electrophysiology benchmark we create) --- we show that MDA sets a new SOTA in terms of data-efficient model learning and reliable interventional forecasting ability.

Many of the series that generative time-series models are benchmarked on place a large probability mass on a single value --- it does not rain, no ride is requested, no part is ordered. We report what happens when such data is evaluated carefully. First, the standard rolling-origin protocol can score a model on a window whose atom structure bears no resemblance to the dataset: on one benchmark the dataset is $42\%$ zeros and the evaluation windows are $13\%$, on another $47\%$ against $5\%$. This is not a cosmetic problem --- it reversed one of our own conclusions, turning the strongest occurrence model in our study into what looked like a cautionary tale. Second, we give a control in which CRPS is invariant \emph{by construction} while the temporal coupling is destroyed, which measures exactly how much that coupling contributes to a chosen statistic. Third, benchmarking seven models on a matched protocol over five seeds, an autoregressive hurdle beats a conditional flow on five of six datasets, by up to a factor of $153$, while the flow's own occurrence statistics vary by up to $62\%$ across training seeds and every baseline is deterministic. Finally, the model ordering is not the same under five different occurrence statistics, and the two that do not share a construction agree with each other least.

Recurrent neural networks (RNNs) are widely used for sequence learning, yet their application is commonly associated with temporal data, although recurrent computation fundamentally operates on ordered sequences rather than on time itself. Building on this observation, we introduce the Ordered Structural Dependency Hypothesis (OSDH), which proposes that multiple admissible orderings of the same observations may reveal complementary structural dependencies inaccessible through a single sequential organization. To operationalize this hypothesis, we propose the Independent Structural Expert Principle (ISEP), whereby projection-specific sequence models are trained independently before their learned representations are integrated through a dedicated fusion model. As a concrete realization, we present Structural Evolution RNNs (SE-RNNs), which employ conventional RNNs as projection-specific structural experts while preserving the underlying recurrent computation unchanged. Proof-of-concept experiments on three synthetic datasets with substantially different levels of structural complexity demonstrate that the proposed architecture consistently benefits from multiple ordered projections when hidden structural dependencies are present, while remaining competitive on simpler datasets. Since OSDH is independent of the underlying sequence-processing model, the proposed framework naturally extends beyond recurrent networks and may be instantiated using alternative architectures. The results suggest a general computational perspective for exploiting complementary ordered representations across diverse structured learning problems.

Adversarial training under long tailed distributions suffers from a dual imbalance: the class imbalance skews the training objective toward head classes, and the adversarial inner maximization may further amplify this bias. Existing methods mitigate this issue by correcting class priors or adapting class wise robust supervision, yet they treat each class in isolation and fail to identify which boundaries drive long tailed collapse. We propose a Confusion Geometry Rebalancing method (CGRm) for long tail adversarial training, a plug in framework that leverages directed robust errors as training signals. CGRm leverages periodic robust evaluations to derive source class loss weights, class wise robust coefficients, and a directed confusion geometry graph. The method then couples feedback weighted robust optimization with graph guided margin correction, thereby boosting the robustness of vulnerable classes and sharpening the critical boundaries that drive long tailed performance degradation. Experiments on long tailed benchmarks show that CGRm achieves consistent robust performance gains over existing methods, with ablations validating the contribution of each component. We provide the code in the supplement.

Federated learning (FL) in Low Earth Orbit (LEO) satellite constellations is affected by non-IID data and irregular ground-station visibility, both driven by orbital geometry. Global aggregation performs poorly when orbit-level class distributions are disjoint, while strong personalisation can be excessive when these distributions overlap. We present FedOrbit, which combines continuous orbit-level training over inter-satellite links, class-aware hierarchical aggregation, quality-weighted feature aggregation with return-rate dampening, and adaptive feature decomposition based on inter-orbit class similarity. Across three remote-sensing benchmarks and two non-IID partitions, FedOrbit achieves the highest accuracy in five of six settings and is within $0.9$ percentage points of the best result in the sixth. The gains over the strongest baseline reach $16.1$ percentage points under Dirichlet partitioning and $8.6$ under pathological partitioning, with the smallest per-orbit accuracy spread in five of six settings.

Autoregressive semantic ID recommenders are constrained by expensive beam-search decoding, which limits the practical length of item identifiers. Parallel generation methods alleviate this bottleneck by predicting all semantic ID tokens simultaneously, enabling longer IDs. However, existing semantic ID methods still rely on manually predefined and homogeneous ID structures, where both the number of semantic slots and the codebook size of each slot are treated as fixed hyperparameters. This ignores the heterogeneous capacity demands of different semantic subspaces and may allocate prediction capacity to slots with limited utility. We show that uniformly expanding semantic slots can provide limited gains, indicating redundant capacity in homogeneous semantic IDs. We propose InforID, a lightweight adaptive semantic target construction framework for parallel generative recommendation. InforID allocates a fixed capacity budget across candidate semantic slots, thereby jointly determining the effective ID length and slot-specific codebook sizes. Experiments demonstrate improved recommendation accuracy under comparable capacity budgets while preserving one-step parallel prediction.

Probabilistic coastal hazard assessments require accurate characterization of tropical cyclone (TC) parameters, yet datasets often contain missing records for the radius of maximum winds (Rmax), a key variable in Joint Probability Method analyses. This study evaluates data-driven approaches for Rmax imputation, including one-dimensional Convolutional Neural Networks (1DCNNs), Long Short-Term Memory (LSTM) networks, and conventional machine learning models. We examine physics-informed input augmentation, temporal modeling, and transfer learning using synthetic RAFT and STORM datasets for pre-training and observational IBTrACS data for fine-tuning. Including the radius of 34-knot winds (R34) substantially improves performance across all model types. Temporal models achieve higher average correlations than non-temporal models despite using approximately an order of magnitude fewer samples, indicating better preservation of relative Rmax variability across storms. This advantage is more pronounced when R34 is unavailable, suggesting temporal information can partially compensate for missing storm-size predictors. Transfer learning does not improve performance, likely because synthetic datasets have lower and less variable Rmax distributions than IBTrACS. These findings demonstrate the potential of temporal deep learning for reconstructing incomplete TC records and highlight the importance of physics-informed inputs, observational data availability, and distributional consistency in coastal hazard assessment.

Background: Deoxyribonucleic acid (DNA) data storage is a paradigm with great potential for ultra-dense and durable information preservation. However, the rapid proliferation of coding schemes, or codecs, each with their own design constraints and reporting practices, has led to a fragmented landscape that lacks a standardized comparative assessment. Methods: We developed an open-source, modular benchmarking platform that systematically integrates and evaluates state-of-the-art DNA storage encoding and decoding methods (codecs). Our approach uses a curated, diverse set of baseline data and applies multidimensional assessment criteria that are aligned with the consensus standard of the DNA Data Storage Alliance. These criteria include encoding/decoding throughput, computational efficiency, error correction performance across substitutions, insertions, and deletions, and cost efficiency. Results: The developed platform integrates standardized wrapper functions for encoding and decoding, allows for the integration of new methods, and automates reproducible evaluations with comprehensive visual and tabular reporting. Benchmarking both contemporary and classical codecs using their default parameters and multiple metrics demonstrates that no single algorithm is optimal across all evaluated dimensions. The trade-offs between information density, success rate, runtime, and cost are quantified and shown to be critical factors in the design of future-proof formats. Conclusions: Our work establishes a rigorously standardized, open-source evaluation framework that enables reproducible benchmarking, supports evidence-based codec selection, and provides the necessary foundation for translating DNA data storage from experimental research into deployable archival systems.

Recent advances in visual generative models have enabled high-quality image and video generation, but evaluating these models often demands sampling hundreds or thousands of images or videos, which is computationally expensive. Existing evaluation methods also rely on rigid pipelines that overlook specific user needs and provide numerical results without clear explanations. Mimicking how humans quickly form impressions of a model's capabilities from only a few samples, we propose the Evaluation Agent framework, which employs human-like strategies for efficient, dynamic, multi-round evaluations, offering detailed, user-tailored analyses. Given a natural-language evaluation request, the agent decomposes it into sub-aspects, generates targeted prompts, samples images or videos from the evaluated model, invokes suitable evaluation tools, and iteratively updates its plan from the observed evidence, covering both predefined benchmark dimensions and open-ended user concerns. The framework is thus efficient, promptable, explainable, and scalable across models and tools. Experiments show that Evaluation Agent reduces evaluation time to 10% of traditional methods while delivering comparable results. We further introduce Open Evaluation Agent (Open-EA) by constructing EA-CoT-10K, a corpus of history-conditioned step-level instruction-tuning records derived from multi-round evaluation rollouts, and training EA-3B from Qwen2.5-3B-Instruct as a local planning backbone that preserves the structured reasoning, tool invocation, and summary protocol of the API-based agent while reducing dependence on proprietary backbones. Experiments validate the API-based agent on established T2I/T2V benchmarks and open-ended queries, and evaluate Open-EA on four in-domain and three out-of-domain T2V generator families, showing partial cross-family transfer of the learned policy.

A key challenge when designing hybrid quantum-classical workflows is the identification of quantum candidates, that is, determining for which specific problems quantum advantages could potentially be achieved. When choosing between several candidates, it is crucial to consider the characteristics specific to the problem, as these can fundamentally determine how successful quantum or classical approaches will be. This implies that specialized expertise is needed to use hybrid quantum-classical workflows successfully. To address this challenge, we propose a domain-specific language (DSL) to express best-practices in solution strategies using a universal representation that is easy to use and share. This DSL provides a flexible approach to design hybrid quantum-classical workflows and to automate decisions for a wide range of problems, supporting decisions down to problem-specific details while remaining technically independent. Furthermore, we propose a framework that is built around our DSL that enables the execution of defined workflows using the ProvideQ toolbox as an orchestration layer. All contributions from this publication are open source.

GUI agents are shifting from metadata-dependent large language models to purely visual multimodal large language models (MLLMs) that operate directly on screenshots. The core task, GUI grounding, requires translating abstract user instructions into precise element coordinates. This task faces a persistent dual obstacle: conventional grounding models lack the semantic richness to interpret abstract instructions, while end-to-end MLLMs suffer from coordinate hallucinations caused by deficient fine-grained perception. We propose a regression-free framework where a frozen MLLM performs instruction parsing and a dedicated grounding model handles precise localization without learning any coordinate regression. A frozen MLLM first elaborates the abstract instruction into a structured visual description rich in layout cues. These descriptions are then fed to a novel Layout-Aware GUI Grounding Model, which performs regression-free localization by matching against layout-prior candidates, inherently suppressing hallucinations and avoiding expensive fine-tuning. The grounding model is trained with only Text/Icon binary labels, requiring no coordinate regression parameters. On ScreenSpot-Pro, our method achieves over 20% improvement in grounding accuracy over end-to-end systems; on Mind2Web, it raises success rate and element selection rate by more than 15%. These results demonstrate that decoupling instruction understanding from layout-aware localization effectively resolves the core challenges of GUI interaction.

Recent advances in learning-based control have enabled impressive achievements in solving complex control problems in various domains. However, since learning-based control may not be able to realize safety-guaranties, it is of great importance to enhance safety and robustness while maintaining good performances. Take vehicle motion \& dynamics control as an example, in order to overcome the pain points of traditional methods such as heavy parameter calibration effort and learning-based control to bring better performance and efficiency in stability \& agility over prior work for state-based vehicle control tasks, in this work, our method aims to develop a curriculum learning controller enhanced with physics-based predictive safety filter. The validation is conducted with the Python-CarSim platform, demonstrating better improvements and scalability under various maneuvers.

Recently, thanks to its ability to extract knowledge from historical datasets, the data-driven approach has been widely used in various phases of the system development life cycle. During real-time system validation, remarkable achievements have been accomplished in developing an intelligent failure analysis based on historical data. However, despite its superiority over other conventional approaches, e.g., model-based and signal-based, the availability of representative datasets persists as a major challenge. Thus, for different engineering applications, new solutions to generate representative faulty data in different forms should be explored. Therefore, in this study, a novel approach based on Hardware-in-the-Loop (HIL) simulation and real-time Fault Injection (FI) method is proposed to generate and collect data samples under single and simultaneous faults for Machine Learning (ML) applications during system validation phases. The developed framework can generate not only sequential data, but also textual data including fault logs. The results show the applicability of the proposed framework in simulating and capturing the system behaviour under faults within the system components.

Reranking medical procedures against patient queries is a critical component of health insurance information retrieval, complicated by a substantial lexical gap between patient language and clinical nomenclature. We present a systematic comparison of two reranking paradigms for this production task: (1) small cross-encoders (MedCPT, MiniLM-L12) fine-tuned with listwise learning-to-rank objectives across layer freezing configurations, and (2) Qwen3-Reranker-4B, a 4B-parameter instruction reranker whose prompt is iteratively refined via an agentic optimization loop driven by GPT-4.1. On a purpose-built dataset of 2,647 queries across 708 insurance services, we find that a 109M-parameter cross-encoder fine-tuned with ListNet outperforms the 4B-parameter model by 2.6 percentage points on NDCG@3 and 13.3 points on Spearman correlation - at 37x fewer parameters. We report practical findings, a scalable LLM based dataset construction pipeline, and deployment trade-offs relevant to production reranking systems. We release our code and a sample dataset to support reproducibility and adaptation to other domains.

AI coding agents now write a growing share of production code, and human security review does not scale at the rate code is generated. The agents in widest use are closed-weight, so a deploying team cannot read their internals. It can instead run an open-weight model as a reviewer over the agent's output. That reviewer's activations are readable. We ask whether reading those activations recovers a security signal that simply asking the same reviewer misses. We fit a single linear probe per model on a corpus of paired vulnerable-and-fixed Python functions, then test it without retraining on real disclosed vulnerabilities whose weakness type the probe never saw in training, across five open-weight reviewer models. On the vulnerabilities fixed by changing a single function, the probe scores the vulnerable function above its fix on 61-67% of cases for every model, beating the 50% chance line. It also beats the same model's prompted YES/NO win-rate read from its logits, under every prompt we try. Asking the model for a written verdict, even with chain-of-thought, returns the same answer on the vulnerable and fixed function most of the time and so cannot tell them apart. Model activations carry a code-security signal that prompting the same model misses.

Theory of Mind (ToM) is essential for agent interactions, yet existing evaluations either rely on static scenarios that oversimplify mental-state reasoning or interactive settings that provide limited diagnostic insight. We present Avalon-ToM-Bench, a fine-grained benchmark that operationalizes ToM through the asymmetric-information mechanics of The Resistance: Avalon. Rather than evaluating end-to-end gameplay, it decomposes ToM into a 2$\times$2 taxonomy -- epistemic versus motivational reasoning crossed with inference versus action -- using human-crafted, perspective-constrained queries. Benchmarking 28 LLMs reveals three insights: 1) Reasoning, not knowledge. Models show strong game-rule comprehension but markedly weaker ToM abilities, isolating failures to social reasoning rather than missing domain knowledge. 2) Expression, not representation. Mechanistic analyses via linear probing and activation steering show that models frequently represent correct mental-state inferences in their hidden states but fail to express them during generation -- linear probes recover 77-82% accuracy versus 62-70% from the models' own chain-of-thought. 3) Policy, not deliberation. Dedicated reasoning training yields substantial improvements whereas test-time chain-of-thought provides only marginal gains (+11.0 versus +1.1 points on average), suggesting that robust ToM depends on a learned reasoning policy rather than increased inference-time deliberation.

Diffusion models have enabled high-quality video generation in recent years, but the high cost of iterative sampling hinders their practical deployment. Few-step distillation alleviates this cost, yet exposes a quality--diversity trade-off between its two dominant paradigms: trajectory-level distillation (e.g., sCM) favors diversity, whereas distribution-level distillation (e.g., DMD) favors quality. Targeting extreme two-step video generation, we introduce DUET, which reconciles the two paradigms through a noise-level duet of experts: an sCM expert takes the high-noise step to lay out diverse structure, and a DMD expert takes the low-noise step to refine appearance detail. Since the two experts are trained independently with their native objectives, DUET sidesteps the optimization difficulties of loss-level combinations and delivers quality and diversity jointly rather than trading one for the other. We further identify the relay interface and the high-noise stage as the remaining bottlenecks, and address them with RL-guided expert adaptation, yielding DUET+. With the Wan2.1-T2V-1.3B backbone, DUET lifts the two-step quality of sCM close to the level of DMD while retaining nearly all of its structural diversity---about twice that of DMD---and DUET+ further improves overall quality while preserving this diversity advantage. Together, these results establish noise-level expert specialization as a simple, effective paradigm for reconciling diversity and quality in two-step video generation.

Accurate 3D neuron segmentation in fluorescence microscopy is critical for neuroscience. However, the sparse and elongated morphology of neurons poses significant challenges to existing segmentation methods. These methods struggle to preserve both local details and global topology, leading to fragmented results. To address this, we propose NeuroRefiner, a multi-agent system that formalizes the human expert workflow involving iterative global observation and local editing. Specifically, NeuroRefiner comprises three collaborative agents dedicated to diagnosing topological errors, generating correction instructions, and validating refinement quality. To facilitate agent instruction-guided segmentation refinement, we propose TopoRefineNet, a dedicated 3D U-Net-based tool that leverages cross-modality feature fusion to generate refined masks. Through multi-round agent reasoning and voxel-level editing, NeuroRefiner produces topologically more accurate segmentations with enhanced interpretability. Experiments on the BigNeuron, CWMBS, and ZBFWB datasets demonstrate that NeuroRefiner outperforms state-of-the-art methods, notably achieving a 3.02% improvement in F1 score on the challenging ZBFWB dataset.

Face presentation attack detection (PAD) aims to reliably detect a wide range of presentation attacks. While PAD methods achieve strong performance within individual datasets, their performance degrades under cross-dataset evaluation. Variations in sensors or lighting conditions can reduce the effectiveness of detectors from near-perfect to nearly random. Foundation models (FMs) have emerged as a promising alternative because typical PAD datasets, such as the MCIO benchmarks (MSU-MFSD, CASIA-FASD, Replay-Attack, and OULU-NPU), are small relative to the scale used for web-based pretraining. However, existing PAD systems primarily focus on CLIP-based foundation models, while overlooking other FMs with different architectures and training procedures. This study addresses this question by systematically evaluating 32 FMs. Zero-shot prompting achieves performance near chance across model families and scales. The vision encoders, when low-rankadapted (LoRA) with fewer than 1% trainable weights, achieve below 2% intra-dataset ACER in most cases, while cross-dataset ACER is substantially higher. LoRA primarily refines the decision boundary within a dataset, suggesting that pretrained representations and the adaptation dataset play a larger role in cross-dataset generalization than the evaluated lightweight adaptation strategy.

Self-evolving agents are usually built around prescribed optimization pipelines: the framework decides how to gather evidence, revise a persistent artifact, select candidates, and stop. We ask whether this task-specific procedure remains necessary when a frontier model acts as the optimizer. We introduce Open-Ended Optimization (OEO), which keeps the objective, permitted interactions, resource budget, data boundary, and evaluation fixed while allowing the optimizer to compose the improvement process online. We compare OEO with two complementary prescribed approaches: SkillOpt, a staged pipeline with bounded edits, and GEPA, a reflective evolutionary search. Across 14 head-to-head comparisons over 8 benchmark-target-model settings, GPT-5.5-driven OEO records 12 wins, 1 tie, and 1 narrow loss of 0.21 percentage points. It uses a median 34.3 percent of SkillOpt's configured target-interaction token budget. A one-shot, zero-interaction control shows that the gains are not explained by a single prior-driven rewrite. However, delegation has a capability boundary: SkillOpt outperforms OEO with a medium optimizer, and a weak optimizer cannot operate through the unchanged OEO interface. In the fully instrumented OEO-SkillOpt pair, trajectory analysis further shows that prescription changes how optimization proceeds more consistently than it changes final behavior. Together, these findings recast prescribed pipelines as capability-dependent scaffolding: essential constraints remain external, but a sufficiently capable optimizer can compose the route from measurable feedback to persistent improvement.

Collision avoidance systems are commonly used to avoid fragmentation events occurring in Low-Earth Orbit (LEO) and Geosynchronous Equatorial Orbit (GEO). However, these events have been growing in frequency as orbital congestion worsens with the launch of megaconstellations. Consequently, conjunction alerts and collision risks are becoming increasingly common. Current practices, which are commonly manual or rule-based, have difficulty scaling to these worsening dynamic environments. To address this intensifying situation, we propose a reinforcement-learning policy for autonomous collision avoidance, trained via Proximal Policy Optimization (PPO) along with an open-source, high-fidelity astrodynamics simulator for training and evaluation. In 1,000 deterministic GEO episodes, our agent achieves a 97.5% collision avoidance success rate, outperforming traditional controllers such as a rule-based baseline (20.7% success) and an impulsive delta-v planner baseline (27.5% success). To achieve these results, we designed a simulator to train and evaluate our agent, using real-world and simulated debris. We simulate Newtonian two-body dynamics using Sun/Moon third-body perturbations, fuel-dependent thrust, and configurable debris fields. The agent is trained with curriculum learning and shaped rewards oriented toward encouraging survival, adequate projected miss distance, and delta-v conservation. Finally, our evaluation consisted of a fully deterministic pipeline, including shared seeds, per-episode logs, and telemetry exports. Our work is a publicly available framework at https://purl.org/sat-trajectory-avoidance

Internal safety scores judge a prompt before any text is generated, and they are validated by how well they separate harmful prompts from benign ones. That separation is then read as evidence that the score will also catch the attacks that succeed. Harmful intent is a property of the prompt. Jailbreak success is an outcome produced later by a particular target model, decoding policy, and judge. A filter tuned on a score that measures the wrong quantity spends its false positive budget on attacks that would have failed anyway. In this paper we audit that inference. Attention based measurements are usually read from prompt dependent locations, so a wrapper changes both the content being judged and the place the signal is taken from. We therefore introduce Active Attention Probing, which supplies a fixed content independent measurement coordinate. We pair every base goal with a plain and a wrapped version and generate real completions from the target models. On Llama, wrapping raises harmful generation from 0.05 to 0.27 while harmful intent AUROC falls from 0.936 to 0.803, so the attacks grow more dangerous while the prompts look safer to the score. Among wrapped harmful prompts the outcome AUROC is 0.220, which places the attacks that succeeded below the attacks that failed. Rare token, passive, and detector derived channels reproduce the reversal on the same matched design, and the reversal itself persists across three target models, seven attack families, and two independent judges. Distribution shift then degrades calibration and threshold transfer before it degrades ranking.

Reliability qualification of advanced semiconductor devices requires sequential stress decisions that balance characterization objectives against multiple competing failure mechanisms. Current practice relies on static test plans derived from population-level acceleration models, which cannot adapt to per-unit variability or real-time degradation observations. This paper presents a closed-loop adaptive test planning framework that formulates reliability qualification as a partially observable sequential decision problem and solves it using Monte Carlo tree search for seed-action simulators (MCTS-SA) coupled with extended Kalman filter (EKF) belief-state estimation. The framework models stochastic, per-device variability in bias temperature instability (BTI), electromigration (EM), and time-dependent dielectric breakdown (TDDB), and treats stress selection as a constrained sequential optimization, i.e., to maximize the probability of successful degradation characterization while respecting catastrophic failure constraints. Under the experimental assumptions used here (discrete stress actions, proxy damage observability, and cumulative degradation without recovery), we believe this to be a novel application of tree-search-based adaptive test planning to multi-mechanism reliability qualification. Across 5,000 planning iterations, the characterization yield (CY) improves from 20% in the first 500 iterations to over 54% in the final 500, with 39% cumulative success, while the best successful test sequence terminates with EM and TDDB damage fractions DEM=0.564 and DTDDB=0.537, well within safety margins. These results demonstrate that sequential Bayesian planning can synthesize damage-aware test policies that significantly outperform non-adaptive strategies for reliability qualification under competing failure modes.

Symbolic regression is the problem of finding an algebraic expression describing a stochastic dependence of a target variable on a set of inputs. Unlike forms of regression that fit parameters assuming a fixed model structure, symbolic regression is a search problem over the space of expressions, represented, for example, as abstract syntax trees using a library of operators. Symbolic regression is typically used in settings with limited, noisy data in the natural sciences. However, searching for a single best-fitting expression fails to capture the epistemic uncertainty about the expression, which motivates a Bayesian perspective that enables uncertainty quantification and specification of natural priors to constrain the search space. In this work, we propose ERRLESS (Entropy-Regularized Reinforcement Learning for Expression Structure Sampling), a scalable approach for sampling from the posterior distribution over expressions given data using maximum-entropy reinforcement learning. ERRLESS learns a neural policy that constructs expressions sequentially by building up their abstract syntax trees. At convergence, the policy samples expressions from the posterior. At test time, expressions can be sampled by rollouts of this policy. We demonstrate that ERRLESS achieves competitive results on the Feynman benchmark while producing short and interpretable expressions. Additionally, we demonstrate that the mean of the posterior predictive approximated by ERRLESS achieves a high coefficient of determination ($R^2$) compared to an SMC baseline, highlighting the benefits of the Bayesian perspective in symbolic regression.

Lane detection requires recovering thin, elongated, and frequently occluded lane structures under challenging driving conditions. While anchor-based detectors provide efficient candidate generation, their performance is limited by two coupled issues: backbone features often lose structural continuity along partially visible lanes, and classification confidence may decouple from line-level localization quality, allowing inaccurate anchors to persist before non-maximum suppression (NMS). We propose a structure-enhanced and quality-aware framework that improves lane representation and dynamic-anchor scoring while preserving the inference pipeline of the Anchor Decomposition Network (ADNet). Specifically, a Gated Horizontal-Vertical Token (GHVT) module enhances mid- and high-level backbone features via lightweight directional token interactions with a learnable residual gate. In parallel, Line-Quality-Aware Dynamic Anchor Scoring (LQAS) calibrates existing classification logits using quality supervision, hard-negative suppression, and pairwise ranking without adding inference branches. On the VIL-100 dataset, our method improves ADNet-R34 from 89.97 to 91.28 in F1 score at the 0.5 intersection-over-union threshold (F1@50), reducing both false positives and false negatives. Additional experiments on CULane and TuSimple datasets, extensive ablations, score-distribution diagnostics, and runtime analysis confirm complementary structural and ranking improvements with minimal computational overhead.

Large Language Models (LLMs) have emerged as powerful tools for improving recommendation systems. The effectiveness of LLMs arises from their ability to harness rich textual information and their capacity to model heterogeneous user preferences based on users' interaction history. However, due to the large-scale and deep architectures, LLM-based sequential recommendation approaches generally incur high inference costs, resulting in a low return on investment. To mitigate this cost, many existing approaches resort to using only the first few tokens of item descriptions, which inadvertently discards valuable information contained in the full text, thereby leading to suboptimal recommendation performance. To address this limitation, we propose a novel Token Selection approach for Preference Optimization in LLM-based sequential Recommendation, i.e., TSPORec, which accurately pinpoints informative tokens throughout the entire textual content to improve recommendation performance. Specifically, we design a three-stage pipeline to select informative tokens and introduce a novel proxy reward to facilitate the implementation. TSPORec not only enhances recommendation performance but also improves computational efficiency. Extensive experiments across two models and datasets demonstrate the superb performance (up to 31.25%) and efficiency (up to 63.4%) of our approach compared with six baseline approaches. Code is available at https://github.com/WNQzhu/TSPORec.git.

Graph Neural Networks (GNNs) suffer from two fundamental limitations: over-smoothing, where node representations become indistinguishable with depth, and over-squashing, where long-range information is compressed through limited message-passing channels. Existing metrics such as Dirichlet energy provide global characterizations of over-smoothing but lack the resolution to analyze node-level behavior and guide architectural improvements. In this paper, we propose LEED (Local Embedding Evolution Distance), a novel local metric that quantifies over-smoothing by tracking the evolution of individual node embeddings across layers. By operating at the node level, LEED enables fine-grained analysis of representation dynamics during training, revealing heterogeneous over-smoothing patterns that are invisible to global energy-based measures. This locality induces informative node importance scores, interpreted as embedding-driven centrality measures. We leverage LEED to design a more efficient strategy for virtual node selection. Unlike existing approaches that depend on multiple heuristic centrality measures, our method uses LEED as a unique criterion to guide the construction of Local Virtual Nodes to mitigate over-squashing. Experiments show that LEED provides more informative diagnostics than Dirichlet energy while preserving global evaluation, and enables more effective virtual node integration, improving GNN performance across datasets.

Compressing large language models to two bits or fewer is increasingly feasible through block-wise post-training quantization; cross-block variants reconstruct neighboring Transformer blocks within a moving window. In the fixed two-block setting studied here, the matched sequential baseline moves this window through the network once, so errors introduced early in the sweep are not revisited. We propose Interleaved Cross-Block Quantization (ICBQ), a scheduling modification that revisits the boundary pair between consecutive chunks. Each seam pair is refined twice: first at the end of one chunk and again at the start of the next. The method retains the local two-block objective and reuses the calibration inputs of existing block-wise PTQ pipelines. Under stated local contraction and smoothness assumptions, we derive a depth-wise upper-bound comparison in which seam revisits multiply the propagated term while the residual remains bounded independently of depth. In the reported experiments, ICBQ reduces ternary-quantization perplexity relative to the matched Sequential CBQ baseline, yields finite perplexity in configurations where the baseline has severe degradation, and can also be used with 3-bit and 2-bit GPTQ.

Recently, AI-driven video generation has attracted considerable attention. This surge increases the demand for reliable video quality assessment (VQA) metrics to evaluate AI-generated content (AIGC) videos and guide model optimization. Existing studies assess video quality through visual harmony, video-text consistency, and domain-specific alignment, yet lack quantitative metrics for measuring fidelity to physical laws. To address this limitation, we present a novel benchmark that evaluates the quality of AIGC videos based on their compliance with physical principles by quantitatively measuring geometric consistency across frames extracted from generated sequences. This serves as a proxy for estimating the extent to which generated videos conform to real-world physical rules. Specifically, GeoCon-Bench captures global motion through translation estimation, fits homography or fundamental matrix models using background correspondences, and reports complementary metrics, including inlier ratio and geometric error. We also release a dataset containing 20 scenes across six motion categories. Experiments on state-of-the-art AIGC models demonstrate the reliability of GeoCon-Bench as a video quality assessment metric.

Recent advances in speech synthesis and audio generation have made high-fidelity acoustic forgery low-cost and difficult to attribute, enabling a realistic attack scenario in which speech and background audio are independently manipulated over otherwise authentic video. Yet existing research either focuses on visual manipulation, addresses speech detection in isolation, or conflates speech and non-speech audio as a single undifferentiated audio stream, overlooking the distinct forensic challenges posed by background audio. This conflation is consequential: the two acoustic components arise from fundamentally different generative mechanisms, exhibit distinct artifact profiles, and pose different challenges to detection systems. We introduce MADBench, the first benchmark that treats speech and environmental audio as distinct acoustic components, enabling component-aware evaluation of audio deepfake detection across independently manipulated forgery sources. We benchmark representative state-of-the-art detectors and multimodal large language models under a unified protocol. Our experiments reveal that environmental audio manipulation is more detectable than synthetic speech across general-purpose encoders, while existing pretrained detectors fail on both acoustic components, and manipulated environmental audio asymmetrically degrades speech deepfake detection, findings entirely invisible under the single-label paradigm of prior benchmarks. MADBench establishes a rigorous foundation for future research into robust, component-aware audio deepfake detection.

Traditional Text-to-SQL research and benchmarks assume a known target database, overlooking settings in which a query must be routed within a large, heterogeneous database collection. We therefore study schema linking in a multi-database setting, where the system must first locate the target database and then construct a compact, SQL-relevant schema for generation. We propose MDB-Link, a hierarchical schema-linking framework that retrieves question-relevant columns from a global index, aggregates retrieval evidence to shortlist databases, and uses a budget-aware large language model (LLM) for database reranking, table selection, and column grounding. With Qwen2.5-14B, MDB-Link outperforms LinkAlign on MMQA, Spider2-Snow, and BIRD-dev in database localization and column selection while producing schema subsets close in size to the gold schemas. Exact match improves from 16.88 to 51.41 on MMQA, 2.50 to 9.17 on Spider2-Snow, and 12.52 to 38.01 on BIRD-dev. MDB-Link also runs faster than LinkAlign and AutoLink, demonstrating the effectiveness of hierarchical schema reduction for downstream SQL generation.

The Independent Chip Model (ICM) converts tournament chips into reference prize equity, and policies are routinely constructed against those values. Because ICM reads only stack sizes, it omits action order, blind obligations, and seat rotation, and it does not price the elimination pressure a big stack puts on the short stacks it can bust. Those omissions can alter the successor-state contrasts that determine a move. We introduce Strategic-Continuation Optimization (SCO), a policy-construction method that enumerates current-hand outcomes, maps them to successor states, prices those states with continuation values computed from the finite tournament model, and optimizes and freezes the resulting current-hand policy. The fixed-ICM comparison policy changes one thing only: the same optimizer solves the same game with successor states priced by analytic ICM, so the two policies differ only through that pricing. We evaluate the resulting policies in a three-player jam/fold tournament with a \$1M prize pool. Relative to the frozen strategic-continuation benchmark, analytic ICM has \$9{,}036 mean absolute value error across all 2,838 state--seat entries. That value error rewrites the ranges it prices: measured against each decision point's own fixed-ICM jam range, SCO moves the jam frequency by an average of 14.08\%. To price those different moves, we compare all 946 states and three policy owners while changing only the focal policy and holding both opponents and the continuation evaluator fixed. The policy produced by SCO earns \$214.33 more prize equity per hand on average and is favored in 2,433 of 2,838 matched units. The ordering survives replacing the solver-built opponent with two LLMs and with a family of non-modeling threshold players. This value-to-policy-to-cost chain shows directly when ICM becomes an inadequate objective for tournament strategy construction.

Cross-Domain Sequential Recommendation (CDSR) aims to alleviate data sparsity by transferring dynamic user interests across related domains. A key challenge lies in effectively bridging these domains. In single-domain modeling, models cannot distinguish between domain-specific and domain-invariant interests. Recent methods merge domain-specific sequences chronologically into a mixed-domain sequence to capture domain-invariant knowledge. However, they typically deploy separate encoders for the mixed-domain sequence and train them with per-domain loss aggregation. This workflow magnifies inter-domain discrepancies and disrupts domain-invariant interest coherence, especially when query target pairs in Seq2Seq originate from different domains. In this paper, we present CoRCi (Cross-Reconstruction for Coherent Interest), a dual-target CDSR framework that tackles these drawbacks. Specifically, CoRCi proposes a Cross-Reconstruction approach that generates mixed-domain representations directly from pre-encoded specific-domain representations via cross-attention. The generated representations are then trained using a single, sequence-level, domain-agnostic loss to preserve the coherence of domain-invariant interests. To further suppress domain discrepancies in mixed-domain modeling, CoRCi introduces FocalNCE, which embeds Focal Loss into the preceding mixed-domain InfoNCE objective. The new loss assigns higher penalties to negatives drawn from the same domain as the query, thereby strengthening domain-invariant alignment. Extensive experiments on four real-world datasets demonstrate that CoRCi consistently outperforms state-of-the-art CDSR counterparts, achieving statistically significant gains across all metrics.

Agent skills, bundles of instructions and resources that an LLM agent loads on demand, form an emerging supply chain where a single poisoned skill can persistently compromise every agent that installs it. However, existing skill attacks either fire on every request or rely on fine-tuned weights or multiple skills, leaving a conditional and low-cost backdoor unexplored. In this work, we present ElasticBack, an effective conditional single-skill backdoor that plants a rule R in the skill document and a benign-looking trigger T in the user query, so the malicious payload fires only when both co-occur. ElasticBack binds the two sides through a trigger-as-switch construction, generating R via semantic-anchored rule injection. It then freezes R and evolves T against it with a stealth-constrained genetic search, so that effectiveness and stealth are optimized, keeping the backdoor weight-free and dormant on benign inputs. Extensive experiments across three target behaviors (50 skills each) and four agent LLMs show that ElasticBack attains a high attack success rate at a near-zero false-positive rate with preserved clean accuracy, transfers across models, and evades deployment-time defenses. These results motivate stronger defenses for the skill supply chain.

LLMs are rapidly embedding themselves into daily life: drafting our emails, managing our schedules, and making decisions on our behalf. As they move from individual tools to participants in multi-agent organizations, an important question arises: do they reproduce the governance failures like free-riding, corruption, and entrenched leadership that plague human institutions? We introduce the Hierarchical Game (HG), a public goods game extended with managerial authority, democratic elections, and private communication. Testing six frontier models across twelve experiments that add institutions one at a time (speech, peers, government, wages, oversight, elections), we find distinct behavioral profiles: Qwen promises and lies (13.3\% broken promises); Grok refuses to cooperate on its own but becomes fully cooperative once a manager can punish it (16\%$\to$100\%); Claude and GPT-4o cooperate reliably at baseline. But honesty proves fragile. When the manager role comes with a salary, all models except GPT-4o start cutting private deals to win or keep the position. When punishment is made anonymous, honest models begin to cheat. When all agents share the same model family, the first elected manager stays in power indefinitely. Leadership change only happens in groups that mix different families.

Hyperbolic geometry has recently emerged as a powerful representation space for multimodal learning, as it naturally captures hierarchical semantic structure across modalities. Despite this progress, how such representations behave under continual learning poses fundamentally different challenges that remain underexplored. This work provides a geometric perspective on this problem and establishes a theoretical foundation for representation preservation in hyperbolic space, showing that preventing forgetting requires cross-modal invariance under a shared hyperbolic isometry. We further show that forgetting in hyperbolic continual learning involves both semantic relation drift and hierarchy-related distortion, motivating preservation of both cross-modal relational structure and hierarchical geometry. Guided by these insights, a principled continual learning framework is derived that preserves essential geometric structure while allowing effective adaptation to new tasks. Experiments on continual multimodal benchmarks corroborate the effectiveness of the proposed approach.

Direct Preference Optimization (DPO) aggregates token-level log-probability ratios via uniform summation, implicitly treating all tokens as contributing equally to the preference signal. However, the contribution of individual tokens to the preference signal varies. We introduce token credit, which modulates each token's KL regularization based on its contribution to the preference outcome. We derive that effective token credit is proportional to the magnitude of each token's implicit reward, and observe that this quantity evolves substantially during training. This implies that static token credit becomes increasingly misaligned as training progresses. In this work, we propose Se-DPO (Self-Evolving Token Credit for DPO), a live mechanism that derives token credit from the model's own evolving internal signals during DPO training. Since the reward signal varies in reliability across positions, Se-DPO calibrates token credit based on both the strength and the confidence of each token's contribution. Se-DPO requires no external models, adding only a lightweight calibration network with minimal computational overhead. Experiments show that Se-DPO improves over DPO by up to 9.8 points on AlpacaEval~2 and 12.2 points on Arena-Hard.

Optimization theory is a widely used tool for intelligent decision-making. While classical optimization deals with fixed, time-invariant objective functions, many modern applications operate in dynamic environments where data arrive sequentially, and the learning objective evolves over time, often under decentralized data and communication constraints. Motivated by these trends, we study decentralized optimization from streaming data through a structured time-varying formulation in which the global objective is a temporally weighted average of losses observed across the network. We analyze multi-iteration decentralized first-order methods, including decentralized gradient descent. For strongly convex and smooth losses, we develop guarantees for the Euclidean-norm \emph{tracking error} through a contraction-mapping viewpoint. The resulting bounds decompose the tracking error into a fixed-point tracking component and a bias term induced by decentralization and data heterogeneity. We specialize our analysis to uniform and exponentially discounted weights, as well as their finite-memory \emph{windowed} counterparts. The bounds explicitly characterize the roles of the temporal weighting rule, per-step iteration budget, step size, and network connectivity. Uniform weighting yields a vanishing fixed-point tracking contribution of order $\mathcal O(1/t)$, whereas discounted and windowed strategies generally induce non-vanishing tracking floors governed by the discount factor and effective memory, respectively. In all cases, decentralization induces an additional non-zero bias floor under a constant step size. Numerical experiments illustrate the predicted trends.

UAV vision-language navigation (UAV-VLN) focuses on enabling an aerial agent to follow natural-language instructions in open 3D environments from egocentric visual observations. Current approaches suffer from three coupled issues: weak grounding of instruction-relevant landmarks in visual observations, insufficient exploitation of long-horizon history, and unstable decisions under local traps or repeated exploration. To address these issues, we propose a unified semantic-to-decision framework. First, we present an instruction-grounded semantic enhancement module that injects object-level semantics and relative spatial cues into the current observation state. Subsequently, we develop a relevance-aware dynamic temporal aggregation strategy that reweights the full history buffer while converting a few high-relevance frames into structured landmark prompts for the decoder. Finally, we devise a topology-aware decision method that combines local-optimum cognition with group-relative policy optimization under progress, goal, semantic, and path-compliance rewards. Experiments on the widely used AerialVLN and OpenFly benchmarks clearly demonstrate that our method achieves state-of-the-art performance.

How expressive is prompting a transformer? Answering this question is important for separating the roles of prompting, architecture, and pretraining in transformer models, and for determining whether task-specific behavior must be stored in model weights or can instead be induced at inference time through the prompt. We show, in an approximation-theoretic sense, that pretraining is optional: a single-layer softmax attention network with random, untrained weights can approximate any Hölder function on a compact manifold when steered by an appropriate soft prompt. Guided by the connection between softmax attention and kernel methods, we construct explicit soft prompts (a prompt per target function, independent of the query) as solutions to linear systems matching attention logits to Gaussian kernel exponents, under which the frozen transformer emulates the classical Nadaraya-Watson kernel estimator. The construction requires only a mild rank condition on the weights, which we show holds almost surely under Gaussian initialization. The prompted network inherits the theoretical guarantees of kernel regression, leading to universal approximation theorems with minimax-optimal rates that depend on the intrinsic dimension. We further quantify the cost of prompting, exposing a tradeoff between the norm of the constructed soft prompt tokens, prompt length, and hidden dimension. Numerical experiments corroborate the constructions and predicted rates.

External natural-language skills provide large language model (LLM) agents with reusable and editable guidance for solving complex tasks. Yet their effectiveness depends not only on skill quality, but also on whether the policy can translate the provided guidance into appropriate actions. However, methods specifically designed to improve this skill-utilization ability remain largely underexplored. In practice, skill-based agents are commonly trained with reinforcement learning objectives centered on task-level rewards, which offer limited supervision and struggle to capture subtle differences in how effectively the policy uses the provided skills. We propose BCSD (Bidirectional Context Self-Distillation), a framework that combines self-distillation with reinforcement learning to train LLM agents to use external skills more effectively. Unlike prior self-distillation methods that rely on a single privileged context, BCSD evaluates each trajectory from two complementary skill-context views. The augmented view introduces higher-level Meta-Skill guidance, while the reduced view prunes general guidance to highlight task-specific skills. Their complementary token-level signals are combined to rescale the RL advantage. Experiments on ALFWorld and WebShop demonstrate that BCSD achieves the strongest overall performance across model scales, enabling agents to utilize external skills more effectively. Ablation studies further verify the complementary contributions of the augmented and reduced context views. Code will be released to ensure full reproducibility.

In the era of large language models (LLMs), attackers often manipulate natural language to elicit unsafe or harmful outputs, creating a new natural language attack surface unique to LLM-based systems, where attacks directly exploit explicit linguistic cues in user prompts to bypass the safety mechanism of LLMs. However, such attacks can often be mitigated by existing safety alignment algorithms. On the other hand, human language is inherently grounded in pragmatics, necessitating typical context to interpret language, e.g., world knowledge, social norms. However, such contexts are often implicit because they are not directly expressed in human language and are not sufficiently leveraged in safety alignment, creating a fundamental mismatch between human language interpretation and safety alignment approaches. In this paper, we demonstrate that this mismatch exposes vulnerabilities in LLMs. We refer to this vulnerability as the pragmatic attack surface, which can be exploited to achieve high attack success rates. The experimental results demonstrate that our proposed approach outperforms baseline attack methods across various open-source and closed-source models by a substantial margin.

Large language models are increasingly deployed in education as tutors, teaching assistants, and content generators. These roles place demands that ordinary question answering does not: a usable education-facing model is supposed to be accurate, safe under sensitive prompts, instructionally useful, and aligned with pedagogical goals at the same time. Existing benchmarks evaluate these requirements largely in isolation, so none assesses education-facing suitability as an integrated profile. We introduce ELBench, the first benchmark to evaluate all four requirements (General Capability, Safety and Trustworthiness, Basic Education, and High-Level Cultivation) on the same models under a common protocol, combining curated public sources with newly synthesized safety and cultivation data. We evaluate nine models, seven frontier general-purpose systems and two education-specialized variants, and report three findings. First, module-level profiles are more informative than a single aggregate: the top six models are statistically indistinguishable on overall score, yet their module leaders differ substantially, and safety is anti-correlated with practical teaching (r = -0.83). Second, the Chinese-developed models lead the safety module, the most discriminative in the suite; this advantage is largest on region-specific normative content and narrows, but does not vanish, on universal-harm content. Third, the two education-specialized models lead neither education module, and on High-Level Cultivation all models share a systematic blind spot: on the structured judgment task they converge on the same non-reference option, favoring pedagogical style over fit to the stated goal, so the module scores uniformly low and does not separate models. This raises, but does not resolve, whether domain post-training keeps pace with frontier systems on education tasks.

Large reasoning models (LRMs) achieve remarkable success on complex tasks but remain vulnerable to harmful prompts that induce unsafe outputs. Recent methods align LRMs using direct refusals or safety rationales, yet often focus on prompt patterns rather than intrinsic attack mechanisms. As a result, these pattern-centric alignments struggle to generalize across diverse jailbreaks, compromising adversarial robustness and reasoning utility. We propose AdvSafe, a dual-adversarial framework that enables LRMs to internalize unsafety knowledge by explicitly deconstructing adversarial mechanisms. This moves beyond pattern-dependent traces, fostering robust cognitive defense without compromising reasoning utility. Our pipeline operates via a two-phase adversarial game. First, in adversarial synthesis, an autonomous agent dynamically crafts deceptive jailbreak prompts, adapting its strategies to breach a strong teacher model. Second, in adversarial extraction, the breached teacher executes a cognitive counter-attack. For every successful jailbreak, the teacher unmasks the camouflage, explaining why the attack succeeds and how such prompts can be identified and mitigated. This dual-adversarial process yields a compact reasoning dataset capturing rich, generalizable unsafety knowledge. Student models trained on this dataset implicitly acquire safety alignment through intrinsic threat comprehension. Experiments show that with only 1K synthesized samples, AdvSafe-aligned LRMs achieve significantly stronger jailbreak robustness than existing baselines, with almost no utility degradation. Furthermore, AdvSafe improves robustness against out-of-distribution prompts, demonstrating that learning unsafety knowledge enables a superior robustness-utility trade-off and generalizes beyond seen attack patterns.

Publicly available Arabic datasets for target-specific stance detection remain limited, particularly for evaluating cross-target generalization. This paper presents the Mawqif-v2 Extension, consisting of 996 manually annotated Arabic tweets collected from three public targets: Women Driving, E-Cars, and Trimester System. Each tweet is annotated with stance, sentiment, and sarcasm labels following the original Mawqif annotation scheme. The released extension is intended as a held-out evaluation set for assessing model generalization to both semantically related and previously unseen targets, while the original Mawqif dataset is used for training and development. In addition, we establish baseline results using several Arabic and multilingual transformer models, as well as zero-shot large language models (LLMs), to facilitate reproducible evaluation. Together with the original Mawqif dataset, the Mawqif-v2 Extension provides a benchmark for evaluating cross-target generalization in Arabic stance detection.

We introduce TCS-Bench, a benchmark for evaluating Large Language Models (LLMs) on research-level Theoretical Computer Science (TCS) proof generation. TCS-Bench consists of theorem-proving tasks from papers published at top theoretical computer science venues (STOC, FOCS, and SODA). Each task provides the necessary context to derive a self-contained proof for a target result. We evaluate state-of-the-art models on this benchmark. We verify the correctness of generated proofs via a verification agent, and further benchmark the verifier against human-expert proof judgements on a set of target statements and generated proofs pairs. Our reference verifier achieves over 90% accuracy on the expert labeled set.

Foundation world models have made remarkable progress in planning, simulation, and embodied intelligence. However, optimizing a pretrained world model toward a user-specified objective remains difficult: each campaign typically rediscovers optimization strategies from scratch, and the resulting knowledge rarely transfers to the next model. Existing research agents automate the optimization loop but treat successful strategies as directly reusable recipes, without principled safeguards for when transfer is appropriate. We argue instead that retrieval is not transfer: a strategy validated on one model is at best an optimization hypothesis for another, and becomes transferable knowledge only after target-side experimental valida- tion. Guided by this principle, we propose VERDI , a continual framework for evidence-licensed world model optimization. VERDI characterizes each world model through shared inference-time probes to construct an Optimization Fin- gerprint, retrieves relevant prior experience as ranked hypotheses, and validates every candidate under a frozen target-side verifier before admitting it as reusable evidence; contradictions among nearby fingerprints further trigger probe evolution, continually refining the diagnostic representation itself. Experiments on Ctrl-World, the Cosmos family, and RoboCoin show that VERDI reduces search cost by 68%, GPU cost by 69%, and negative transfer from 0.34 to 0.06, while predicting transfer outcomes with 83% sign accuracy.

Enterprises increasingly seek to query data lakes using natural language via AI-driven tools like semantic operators or deep research agents. However, the latter operates as an opaque black box, hiding its intermediate reasoning and data retrieval steps, and failing to expose controls for managing API costs and execution latency. Meanwhile, the former can be prohibitively expensive for enterprise-scale data lakes. Consequently, analysts using these systems lack the agency to intercept hallucinated premises, verify intermediate results, or correct the system's trajectory. We present Carnot, an interactive execution engine for AI-driven analytics. Carnot compiles natural language requests into physical execution graphs and surfaces them through an interactive notebook interface. Rather than waiting blindly for a final output, users can critique the plan, incrementally execute operators, inspect intermediate data, or directly edit the underlying code or semantic operator instructions. Carnot's query optimizer will optimize the query with respect to cost or latency constraints provided by the user. Our demo will showcase how Carnot helps users achieve efficient and verifiable insights on workloads motivated by real enterprise use cases.

Compiler optimizations are essential for achieving high performance in modern software. However, recent studies highlight the persistence of performance bugs, i.e., subtle defects where the compiler generates functionally correct but computationally inefficient code, leading to significant performance degradation. Existing detection and testing methods typically employ a bottom-up approach, focusing on specific low-level code properties and remaining confined to known optimization rules. Consequently, they struggle to quantify the holistic impact of identified issues and often overlook critical microarchitectural inefficiencies. We observe a key indicator of untapped potential: different compilers often produce binaries with significant performance differences for identical source code. However, the root causes of these discrepancies remain largely unexplored and difficult to pinpoint using current techniques. To bridge this gap, we introduce a top-down differential analysis methodology. This approach calibrates compiler optimization differences with fine-grained, hierarchical microarchitectural metrics, offering a comprehensive view of runtime behavior. Using a sampling-based approach, this method efficiently pinpoints the critical code snippets responsible for performance differences, enabling targeted root cause analysis. Our empirical evaluation uncovers substantial and often surprising performance differences between binaries generated by GCC and Clang. A categorization of root causes reveals systemic challenges in compiler optimizations. To quantitatively validate our findings and demonstrate practical impact, we developed a binary patching framework that fixes identified performance issues by transplanting superior code sequences from competing compilers. This work provides a novel lens for understanding and analyzing optimization defects.

Real-world cyberattacks often require sustained progress across multiple hosts and network segments, making multi-hop cyber ranges essential infrastructure for studying and improving LLM agents' ability to sustain complete attack chains. Prior work has scaled isolated vulnerability tasks and constructed multi-host scenarios from manually specified vulnerability semantics. However, they are still unable to automatically orchestrate the growing supply of vulnerability environments into end-to-end validated multi-hop ranges. To this end, we present RangeFactory, an automated cyber-range orchestration framework that constructs multi-hop cyber ranges at scale from isolated vulnerability environments. RangeFactory formulates range construction as dependency resolution: it extracts dependency information from agents' actual attacks against real vulnerabilities, resolves known dependencies through template-guided orchestration, and uses end-to-end attack execution to validate runtime dependencies that emerge after composition. Using RangeFactory, we construct RangeBench with 1,148 validated range instances spanning 287 distinct attack chains and evaluate frontier attack agents across attack depth, network scale, and task information. Among runs that compromise the entry vulnerability, 24.5-47.0% still fail to complete the remaining attack path, revealing a substantial sustained-compromise gap between establishing an initial foothold and completing a multi-hop attack. RangeFactory further produces a corpus of 5,541 outcome-annotated multi-hop attack trajectories, providing execution data for attack-process analysis and future agent training.

Incident response planning is critical for restoring compromised software systems after cyberattacks. Common practice relies on expert-driven playbooks that encode fixed response procedures, but these static workflows struggle to adapt to evolving incident states, changing recovery objectives, and execution feedback. Recent LLM-based planners and tool-using agents improve automation, yet they remain unstable in long-horizon response because they lack a unified basis for maintaining incident state, aligning actions with the current recovery stage, and reusing historical experience. We present STAIR, an end-to-end agentic planning framework for incident response. The framework maintains the current incident as Graph-as-State, uses a Stage Router to dispatch planning to stage-specialized agents, and retrieves historical experiences to guide action selection. An Execution Harness executes actions, returns feedback to update the incident state, and validates action effects for future experience reuse. Across 100 Docker-based cyber ranges, our framework achieves a normalized defense score of 0.94 and improves over the strongest baseline by 9.5%.

Deep neural network (DNN) training with stochastic gradient descent (SGD) and its variants achieves strong empirical performance, yet classical optimization theory does not fully explain this success. This limitation arises because conventional analyses rely on assumptions such as differentiability, convexity, or smoothness, which are often violated by DNN objectives. In this paper, we establish a unified optimization framework for DNN training by generalizing classical convexity and smoothness through Legendre functions and convex conjugation. Specifically, we introduce $\mathcal{H}(ψ)$-convexity and $\mathcal{H}(Ψ)$-smoothness, which unify convex and non-convex as well as smooth and non-smooth objectives within a single formalism and reveal a natural duality between generalized smoothness and convexity. Building on these generalized properties, we introduce generalized gradient descent (GD) and generalized SGD through convex conjugation. We theoretically prove that generalized GD admits an optimal learning rate of exactly $1$, and derive rigorous gradient-energy-based convergence rates for both proposed optimizers. We further reformulate DNN training as a composite optimization problem, demonstrating that its convergence relies on jointly reducing the gradient energy and controlling the induced norm of the network Jacobian. To characterize the practical influences of network architectures and training configurations, we introduce the gradient correlation factor and model capacity risk, and quantitatively analyze how architectural designs, batch size, and model capacity shape training convergence. Extensive experiments across diverse network architectures, datasets, optimizers, and loss functions validate our theoretical bounds and demonstrate precise alignment between our theoretical predictions and empirical training dynamics.

Activation-based tools are usually tied to one model's native hidden space, requiring probes, sparse autoencoders, and natural-language interpreters to be rebuilt or rediscovered for each new language model. We present a Universal Activation Bus, a framework that provides a common activation interface across compatible language models. Using a small set of source models, we learn a shared dense space together with one lightweight linear encoder--decoder adapter pair per model. After source training, the interface is frozen; a new model joins by fitting only its adapter pair on unlabeled matched text. The resulting interface allows activation-based tools to be shared across connected models, including common probes and SAE features as well as access to an NLA originally trained for a different model. Across five models, semantically related texts form consistent neighborhoods in the shared space, and an onboarded model reuses these tools effectively without retraining them. We further show that an intermediate activation from one model can be used by another model's frozen upper layers to produce predictions. These results establish a stable, model-wise activation contract for reusable tools across compatible language models.

We present a reproducibility study of XFeat, a lightweight local feature extractor and matcher designed to identify corresponding points across images efficiently on resource-constrained hardware. We re-implement the architecture based on the paper and supplementary material, re-evaluate the authors' released checkpoint alongside our re-implementation, and conduct additional architectural ablations to examine design choices that were not fully justified in the original work. This distinction between re-evaluation and reproduction is important, as the paper, supplement, and public code differ in several implementation details, including the backbone layout, fusion block, and training losses. Empirically, our reproduced models closely match and, in some cases, outperform the re-evaluated original checkpoint on MegaDepth-1500 and ScanNet-1500, supporting the main claim that XFeat provides a strong accuracy-efficiency trade-off for standard image-matching benchmarks. Our ablations provide a more nuanced view of two architectural arguments from the original paper. In particular, the parallel keypoint branch is important for semi-dense matching, but its benefit is less pronounced than originally claimed, while the evidence for the specific placement of the single skip-connection remains inconclusive. Finally, we reproduce the original downstream evaluations and find close agreement for homography estimation, while Aachen visual localization remains below the reported results, even for the released checkpoint, suggesting sensitivity to underspecified evaluation details. We then extend the analysis to zero-shot out-of-distribution and cross-modal matching across retinal, thermal-visible, and multimodal remote-sensing imagery, where XFeat remains effective in some settings but degrades sharply under severe modality shifts.

Active inference offers a unified framework for perception, learning, and action, but scaling discrete active-inference models to rich spatial and temporal domains remains difficult. Renormalising generative models (RGMs) address this challenge by composing discrete generative models across spatial and temporal scales, coarse-graining lower-level states and paths into higher-level causes for objects, events, and action. However, fully reproducing and adapting the framework remains difficult: the mathematical exposition is compact, and the reference implementations are deeply integrated within specialized software environments, leaving many algorithmic details implicit. This paper addresses these challenges by providing a self-contained, derivation-oriented account of RGMs together with an open, verified implementation. We explain how the hierarchy is built, how beliefs and actions are updated within it, and how information is passed between levels. Where the published equations and implementation differ in emphasis, we make those choices explicit and explain their modelling consequences. By clarifying the theory and separating it from its original implementation context, this work lowers practical barriers to entry and makes RGMs more transparent, auditable, and reproducible, providing a foundation for future quantitative evaluation and development on machine-learning benchmarks.

Detecting machine-generated disinformation on social media is increasingly difficult as large language models (LLMs) make it easier to generate and rewrite misleading content at scale. Static benchmark evaluations, measuring detector performance on fixed held-out datasets, do not capture how detectors behave when posts are deliberately transformed to evade classification. This paper adapts the Build it, Break it, Fix it framework into Build it, Break it, Repeat (BiBiR): iterative sessions designed to stress-test detectors' robustness under iterative adversarial conditions, evaluating whether models remain reliable when disinformation posts are systematically transformed to evade classification. Across five iterations, the findings show that the best adversarial breakers' transformations came from a combination of back-translation and LLM persona-based rewriting, with the best performing technique achieving a 95% label flip rate (LFR), whilst still preserving the meaning of the original posts. The best builders' model was a triplet contrastive model with a dynamic anchor switching (DASS) architecture, which achieved an average accuracy of 72.68%, outperforming the strong baseline (a fine-tuned e5-small-LoRA) by 15 percentage points on the most robust set of breakers' adversarial attacks. The results demonstrate that an iterative framework best exposes detector weaknesses and pushes robustness improvements; however, it may still require semantic preservation analysis to distinguish valid adversarial evasion from transformations that changed the original disinformation claims' meaning.

Natural language user preferences provide an interpretable interface for LLM personalization. However, universal preference summaries often contain information irrelevant to a particular downstream task. Directly supplying the full preference summary therefore wastes context capacity and introduces cross-task distraction, while manually designing task-specific preference views is difficult to scale. In this work, we study \emph{task-specific preference adaptation}: given a universal user preference summary and a downstream task, derive a task-conditioned representation that preserves sufficient decision-relevant evidence while removing redundant context. To this end, we propose \textsc{AlignXada}, a training-free meta-learning framework that induces reusable textual refinement policies for adapting universal preference summaries to task-specific ones. The refinement policy is iteratively optimized by a meta learner through verbal reinforcement learning. Across 13 tasks and three downstream models (39 task--model cells), \textsc{AlignXada} achieves an average gain of 3.82 points, improving 33 cells while retaining only 22.8\% of the original profile tokens and outperforming RAG in 36 cells. An extended faithfulness analysis further shows that the refined profiles remain largely grounded in the source preferences while preserving task-relevant personalization signals, suggesting that profile-side adaptation serves as a practical complement to universal memory construction for lifelong personalized agents.

We consider the extensive-form bandit problem where on each trial the learner plays an extensive-form game against an oblivious adversary. We focus on the notion of switching regret, which measures the expected performance of the learner against that of any switching sequence of mixed strategies in retrospect. Our algorithm takes a parameter $ρ>0$ and achieves a switching regret of $\tilde{\mathcal{O}}((1/ρ+ρK)\sqrt{H A T})$ where $K$ is the number of switches in the comparator sequence, $H$ is the maximum number of the learner's information sets that can be traversed during a play of the game and $A$ is the number of actions that the learner can possibly take. Our algorithm is extremely efficient, taking a per trial time of only $\mathcal{O}(H B)$ where $B$ is the maximum number of actions available to the learner at any of its information sets.

In this paper we provide Monte Carlo and deep neural network approximations for stochastic representations of solutions to linear elliptic partial differential equations with constant diffusion, drift and killing. Building on the modified Walk-on-Spheres algorithm of Beznea et al. (arXiv:2209.01432), we introduce Monte Carlo estimators that explicitly incorporate sampled random times arising in the analyzed stochastic representations. We establish uniform error bounds for these estimators and show that, under suitable assumptions, a prescribed approximation accuracy is achieved with sample complexities growing at most polynomially in both the inverse accuracy and the problem dimension. Furthermore, we prove a deep neural network approximation result for the stochastic representations. Assuming suitable neural network representations of the boundary data and the distance function to the boundary, we use the constructed Monte Carlo to design deep neural networks that approximate the representation uniformly with a number of parameters growing at most polynomially in the inverse accuracy and the problem dimension. These results extend previous complexity analyses to a broader class of elliptic equations involving drift and killing.

Task arithmetic treats fine-tuning displacements as composable directions in weight space, yet it remains unclear when parameter addition reflects predictable changes in model function. We separate parameter geometry from functional geometry and measure pairwise functional non-additivity over a two-dimensional task-vector surface, using a first-token predictive-distribution interaction ratio conditioned on an input distribution and evaluated with norm-matched controls, three training seeds, and response-only fine-tuning. On Qwen2.5-1.5B, code+safety is more non-additive than the matched code+math control on code and instruction prompts, but not on math prompts. In a prospectively specified six-task expansion, all eight high-versus-low comparisons of unseen task pairs have the predicted sign. The primary ordering further persists under full-parameter fine-tuning at 0.5B, Qwen2.5 LoRA scale tests up to 7B, and a Llama-3.1-8B cross-architecture audit. External validation exposes a sharper boundary: raw public code, instruction, and safety prompts preserve the continuous contrast, whereas an instruction-style wrapper collapses it on the identical public-code prompts, and EvalPlus pass@1 interactions do not robustly reproduce it. Weight-space composition therefore supports coarse, input- and format-conditioned functional statements across adaptation methods, scales, and one additional model family, not a universal merging-performance predictor.

Cooperative capabilities in language models are dual-use. The same social reasoning that supports civic deliberation can also enable strategic omission, false consensus, and manipulative framing. We argue that Cooperative AI evaluations should separate what models can do under benign instructions from what they tend to do under realistic civic pressure. We introduce DiffCoop-Civic, a 10-scenario pilot evaluation suite spanning preference understanding, evidence and persuasion, commitment design, asymmetric information, and dissent preservation. Across seven models from four model families, subtle omission pressure produces a near-uniform shift: manipulative enablement rises by 1.17 points and dissent preservation falls by 1.67 points on a 5-point scale. Overt false-consensus pressure behaves differently: it triggers refusal or redirection in some aligned API models, but direct compliance in several open-weight models. A lightweight Pareto-Trace prompting intervention improves pressure robustness without simply relying on hard refusal. An anonymous reproducibility package is available at https://anonymous.4open.science/r/diffcoop-civil-771C.

Recently, tensor-based physics-informed neural networks (T-PINNs) have received increasing attention. However, existing T-PINNs still face a fundamental challenge: they mainly rely on pre-specified low-rank tensor decompositions with manually tuned ranks, which limits their ability to capture the underlying structures of multivariate solution functions and hinders their practical deployment. To address this challenge, we propose a hierarchical rank-evolving (abbreviated as HRE) representation for multivariate functions, which endows us to faithfully capture the underlying structure of the targeted multivariate function accompanying with automatic rank determination. Concretely, in the hierarchical design of HRE representation, the target multivariate function is decomposed as a small-scale inner tensor with a set of univariate functions along each mode, where a customized tensor network decomposition can be readily deployed to capture the underlying structure of the small-scale inner tensor. In HRE representation, the crucial hyperparameters, ranks, can be adaptively revealed during the decomposition, freeing us from manual rank tuning and making HRE practically applicable to real-world problems. Besides, we build the HRE-PINNs correspondingly. Extensive numerical experiments, including high-dimensional static problems (Helmholtz equation and Poisson equation), nonlinear time-dependent problems (Klein-Gordon equation), and complex fluid-dynamics problems (flow mixing equation and Navier-Stokes equation), demonstrate that HRE-PINNs consistently outperform existing state-of-the-art approaches in terms of accuracy.

All-in-one image restoration is a unified low-level vision task that aims to effectively recover high-quality images from inputs degraded by various types and levels of corruption using a single model. Recent works have achieved remarkable progress by learning degradation-adaptive prompts or network architectures. However, these methods typically apply a uniform restoration strategy across the entire image, neglecting the fact that different regions may suffer from distinct degradation types and varying degrees of severity. In contrast, we propose to perform restoration at the pixel level, thereby enabling more fine-grained and precise control over the restoration process. Specifically, we present MGN-AIR, a novel pixel-level restoration framework for all-in-one image restoration. Our approach first learns to estimate a pixel-level visual prompt. Then, it leverages both textual and visual prompts to provide global and local degradation cues, guiding the model on where to look and how to restore at each pixel. We conduct extensive experiments on multiple all-in-one image restoration benchmarks, covering a wide range of tasks including denoising, deraining, deblurring, dehazing, desnowing, and low-light enhancement. Experimental results demonstrate that our proposed method consistently and significantly outperforms existing approaches.

A model alone does not determine how a programming agent acts. What the model sees, how actions enter the environment, how feedback returns, and how one run affects the next all depend on how the harness is organized. Minimal examples usually show only the basic interaction between a model and tools, while production systems spread these relationships across complex components and dependencies. This paper studies a compact harness design by following a single request through context formation, model decision, environmental action, observation return, and state continuation. Three boundaries---model, execution, and state---connect the model service, tool environment, and persistent state, while the request lifecycle determines the order in which these transitions occur. Together, they show the harness's core role: turning model generations into environmental actions, carrying runtime feedback into later decisions, and allowing state to continue across requests. On top of this runtime structure, a harness can also be gradually refined across runs through continued bootstrapping. The design is realized in the executable artifact https://github.com/lilinxi/Coderlet.

Code smells significantly threaten software maintainability by accumulating technical debt, yet developers often lack the resources to manually address these flaws under tight release schedules. While static analysis tools like SonarQube provide precise detection, they function largely as passive alert systems, leaving the burden of refactoring on developers. To bridge this gap, we present a novel cleaning tool, namely SmellCC, a Visual Studio Code extension that augments SonarQube with an LLM-based pipeline to automatically detect and refactor Python code smells. By employing Chain-of-Thought (CoT) and few-shot learning, SmellCC provides in-place, one-click remediation for the top-10 most frequent smells, effectively preventing the accumulation of technical debt during development. Our quantitative evaluation demonstrates that our SmellCC is promising in helping developers effectively eliminate code smells (96.8\% cleaning rate) with high accuracy (i.e., 91.3\%), ensuring that the refactored code remains syntactically correct and behavior-preserving, thereby significantly improving long-term software maintainability.

Cowork agents may complete benign tasks while disclosing protected data, manipulating unauthorized state, invocate unauthorized API. We define behavioral safety and introduce ActBench, a self-evolving benchmark that evaluates such behavior risk from execution trajectories rather than final responses. Each case pairs a benign task with an adversarial variant that preserves its instruction, configuration, initial state, rating model, and trusted records while injecting a task-reachable payload. ActBench contains 600 cases from 213 scenarios, spanning 15 risk behaviors, six execution spaces, and 48 web-service APIs.To move beyond static payloads, we propose a reward-guided beam search method that jointly optimizes attack effectiveness and task utility, while reflection diagnoses failed execution checkpoint and guides payload revision. Besides, we propose a dual evidence verification mechanism that verifies agent execution safety and utility through log evidence and LLM-based trajectory evidence.We evaluate 15 LLMs and 6 open-source cowork agents over 24,000 trajectories. Under a fixed harness, attack success rates ranges from 10.1% to 94.4% across models, while under a fixed base model, they range from 73.7% to 94.4% across agents.These results show greater variation across models than agent harness, while attacks remain highly successful across all tested harnesses.Our benchmark is released at: https://github.com/zjuicsr/ActBench.

State Space Models (SSMs), as a mainstream research direction of linear Transformers, aim to achieve higher efficiency than standard Transformers in long-context modeling. However, existing SSMs suffer from limited input adaptivity and constrained memory capacity, leading to information loss when modeling ultra-long sequences. To address these limitations, we propose MixFormer, a novel linear Transformer that integrates a Mixture-of-Memory-Experts (MoE) mechanism. Specifically, the model maintains differentiated memory states through multiple collaborating memory experts and employs a novel Time-Aware Linear Attention (TALA) mechanism, which leverages learnable exponential decay functions and positional biases to dynamically update memory. This design enables the model to selectively reinforce important historical information while effectively mitigating memory dilution, substantially improving long-range dependency modeling. Experiments on long-sequence text and image generation tasks demonstrate that MixFormer not only achieves significant performance gains but also provides a more sustainable computational backbone for the next generation of web infrastructure.

Unmanned aerial vehicle vision-language navigation (UAV-VLN) requires agents to translate visual observations and language instructions into reliable flight actions in complex environments. Although recent end-to-end UAV vision-language-action (UAV-VLA) policies reduce reliance on separately designed perception, planning, and control modules, their behavior-cloning objectives provide limited corrective supervision for interactive closed-loop execution. Reinforcement learning (RL) offers a promising solution, while its effectiveness is constrained by inefficient use of samples, long-tailed scene distributions, and policy distribution shift during optimization. To this end, we propose RecoverFly, a failure-aware RL post-training framework for end-to-end UAV-VLA policies. Specifically, RecoverFly adapts token-level RL for stable optimization of grammar-constrained autoregressive UAV actions, revisits unresolved failure cases to strengthen corrective learning and sample utilization, and combines a two-stage long-tail scene curriculum with reference-policy regularization to improve scene adaptation while preserving acquired capabilities. Experiments on the TravelUAV benchmark demonstrate that RecoverFly achieves the best performance on the seen, unseen-map, and unseen-object splits. Moreover, compared to the AerialVLA initialization, RecoverFly improves success rate by 3.12 to 8.37 percentage points under a total rollout budget of about 30\% of the training-set size, validating its effectiveness, robustness, and generalization capabilities.

Cardiac digital twin research is moving from subject-specific anatomical replicas toward virtual cohorts that represent clinically relevant population subgroups. Yet access to representative imaging-derived anatomy datasets remains limited by cohort size, subgroup sparsity, and data-sharing constraints. Conditional generative models could help address this gap, but virtual cohorts are useful only if they preserve realistic, metadata-dependent anatomical variability. Existing cardiac anatomy generators largely rely on conditional variational autoencoders (cVAEs), which couple representation learning and metadata conditioning through a shared regularized latent prior. We introduce CAN-FLOW, a two-step Conditional ANatomy generation framework based on normalizing FLOWs that first learns geometry-only latent representations of diffeomorphic cardiac shape momenta and then models their sex-, age-, and body-mass-index-dependent distribution with a conditional normalizing flow. We trained CAN-FLOW on 2,208 healthy UK Biobank subjects and compared it with cVAEs across regularization strengths. CAN-FLOW generated plausible stochastic biventricular anatomies that better reproduced clinical phenotype distributions, metadata-dependent trends, subgroup variability, point-cloud coverage, and high-dimensional shape variability. Together, these results establish CAN-FLOW as a shareable framework for generating realistic, stochastically varying, metadata-conditioned biventricular anatomies for virtual cohort construction and in silico clinical trial workflows.

On--off cycling is the main cause of compressor wear in residential heat pumps, yet reinforcement learning (RL) controllers for buildings typically optimise only energy cost and thermal comfort, ignoring how much the learned policy cycles. We add a levelised compressor-wear term to the control reward and study how the resulting behaviour depends on the RL algorithm. Training Soft Actor---Critic (SAC) and Proximal Policy Optimisation (PPO) on an identical Markov decision process for the BOPTEST bestest hydronic heat pump case, we find that SAC learns a continuous modulation policy that keeps the compressor permanently engaged---the operating principle of an inverter-driven heat pump---achieving zero start-ups per day, whereas PPO collapses to bang-bang control that cycles more than the baseline. On the BOPTEST emulator the SAC policy cuts thermal discomfort by up to 90.7% for an 11.5% cost increase, while eliminating all baseline cycling.

Betting-based sequential tests and Blackwell approachability are linked by a rate-explicit reduction through support-function residuals. For a compact convex target $S$ and vector observations $r_t$, an OCO learner selects a predictable normal $w_t$ and produces $q_t=\langle w_t,r_t\rangle-h_S(w_t)$. We prove the exact pathwise identity $$ \dist(\bar r_T,S) =\frac1T\sum_{t=1}^Tq_t+\frac{\Reg_T}{T}. $$ When $|q_t|\leq B$, composing this identity with one-sided betting yields a finite-time transfer: if the OCO and log-wealth regrets are at most $a_T$ and $\ell_T$, respectively, then a target gap exceeding \[ \frac{a_T}{T} +2B\sqrt{\frac{\log(1/α)+\ell_T}{T}} \] forces rejection by time $T$, while non-rejection certifies the converse radius. We then formulate a controlled stochastic experiment in which an action selected after $w_t$ satisfies Blackwell's supporting-halfspace condition for every null mean payoff. The resulting wealth is an e-process under adaptive nulls; sublinear OCO regret gives stochastic approachability, whereas persistent mean separation under an alternative gives exponential wealth at rate at least $δ^2/(4B^2)$. Deterministic Blackwell games and passive tests are, respectively, the noise-free and singleton-action cases of this protocol. Bounded two-sample means, kernel MMD, and active heterogeneous data sources instantiate the reduction. The resulting connection is exact algebraically, quantitative at finite time, and operational when experiments are controlled.

On-policy distillation (OPD) aligns a student with a teacher on trajectories sampled from the student itself, reducing the train-test state mismatch of offline distillation. The same feedback loop can nevertheless be unstable: each update changes both the policy and the states on which the next update is computed. We introduce WDL-OPD, a mixture-constrained co-training method with two trainable policies. An anchor policy generates every rollout, an auxiliary policy evaluates the same visited states, and a geometric mixture of their token distributions is matched to a frozen teacher by reverse KL. Both policies receive gradient. We show that freezing the auxiliary recovers an anchor-plus-contrast proxy target closely related to OPD$^2$ and W2S-OPD, whereas joint training creates branch-level degrees of freedom that a static delta cannot express. In recorded Qwen3 experiments at 1.7B and 4B scale, WDL-OPD produces the strongest student checkpoint in each of four scale-domain settings. It raises MATH500 accuracy from 0.630 to 0.685 at 4B and from 0.521 to 0.585 at 1.7B. In code generation, seven single-policy OPD configurations exhibit entropy growth or trajectory degradation, while co-training reaches independently re-evaluated development scores of 0.637 and 0.375. Because several comparisons differ in curriculum or initialization, these results support a stabilization hypothesis rather than a universal causal claim. We provide the exact training algorithm, failure evidence, and the controlled comparison matrix needed to test that hypothesis.

A main promise of looped language models (LMs) is depth-adaptive inference. By iterating a block of shared layers a variable number of times, the model can use less compute for "easy" tokens and more for "hard" ones. However, this adaptivity breaks standard batching: tokens in the same batch now require a different number of loops, so there is no unified forward pass, making efficient inference difficult. Standard inference frameworks like vLLM schedule on the token level and cannot handle this because tokens need to be removed from the batch within the forward pass. Loop-level scheduling has been proposed as a solution, but never implemented end to end. The key challenge is that looped architectures also contain non-looped boundary stages (e.g., token embedding and LM head) that must be scheduled at different frequencies than the loop. We introduce continuous depth batching (CDB), which schedules at the granularity of individual loop iterations. CDB handles boundary stages and loop steps in separate priority queues, makes exit decisions one step ahead, and overlaps all scheduling work with GPU computation. On Ouro 1.4B and Huginn 3.5B, CDB can realize up to $99\%$ of the theoretical maximum speed-up from adaptive-depth, translating to $1.5$-$1.9\times$ higher offline throughput and $45$-$90\%$ lower normalized latency under dynamic serving load.

Large language model (LLM) agents can support medication review between clinical visits, but safe choices for older adults with multimorbidity depend on conditions, medications, and geriatric risks that users may omit. We introduce ATLAS, a coupled graph--policy distillation framework for patient-adaptive medication safety. ATLAS structures guideline evidence as a medication-safety graph. Targeted questions update the patient state and distill relevant relations into a patient-specific medication conflict graph (PMCG). A risk-first multi-agent policy uses the PMCG to screen contraindications, assess cautions and monitoring needs, identify safer alternatives, and verify the final medication plan. We also introduce GeriMedBench, an interactive benchmark that tests safety-critical information acquisition and evidence-based decision revision. Across a European non-interactive multimorbidity benchmark, an Asian interactive multimorbidity benchmark, and an Asian non-interactive cross-guideline benchmark, ATLAS achieves the strongest complete-decision performance among the compared systems. On the European non-interactive multimorbidity benchmark, it exceeds the strongest proprietary LLM baseline by 53.73 points in Strict Success Rate and 14.63 points in overall safety reasoning score (OSRS), with no unsafe recommendations under the automated evaluator. A blinded clinician evaluation gives ATLAS higher mean ratings across all five criteria and flags potentially unsafe recommendations in one ATLAS case and two Gemini cases.

Understanding dynamic sound sources requires jointly determining what produces a sound, where the source is located, and how it moves over time. Yet existing audio-language models often represent clips as global acoustic events, while vision-language models lack the spatial audio cues needed to localize and track individual sources. To evaluate this missing capability, we introduce ST-OmniQA, a spatio-temporal audio-visual question-answering benchmark built from panoramic videos paired with synchronized first-order Ambisonics (FOA) audio of moving sound sources. It contains 40K videos and 400K question-answer pairs organized into four capability levels covering sound-event recognition, direction of arrival, source distance, motion trajectories, and temporally grounded audio-visual reasoning. Building on this benchmark, we propose ST-Omni-R1, which integrates FOA-derived semantic and trajectory representations with panoramic visual context and is trained through progressive curriculum learning and reasoning-tree reinforcement learning. ST-Omni-R1 achieves 77.83\% average semantic accuracy across the four levels, compared with 37.28\% for the best evaluated baseline. Results on three public spatial-audio benchmarks further indicate that its learned spatial and motion representations transfer beyond ST-OmniQA.

In regulated domains such as finance, a model that cannot be explained cannot be deployed, yet many interpretable classifiers defeat their own purpose by producing formulas with dozens of features that no regulator could read. We take the reverse direction. Starting from an interpretable classifier expressed as a single equation over the input features, we progressively simplify it into more readable forms, including a pruned monomial, a directional if--then rule, and the integer scorecards and tallies that finance already deploys. Because the equation is itself the predictive model rather than a post-hoc explanation we can directly quantify what is lost under each simplification. Across four financial datasets, we find that pruning is nearly free and that fidelity can erode faster than predictive performance, allowing simpler rules to remain effective classifiers without faithfully reproducing the original model. A human assessment shows that simplification improves perceived readability, while preferences for different representations vary by professional background. Beyond measuring these losses empirically, we show that some can be anticipated from the original model: we derive a bound on the change caused by pruning and predict how faithfully a rule retaining only the direction of each feature's effect preserves the original ranking.

Transformer-based language models rely on self-attention, whose computation is permutation-equivariant and therefore lacks an intrinsic mechanism for representing token order. Existing architectures address this limitation by explicitly incorporating positional information through learned positional embeddings or hand-crafted positional encodings, such as rotary positional encoding (RoPE), treating positional information as an architecturally acquired capability rather than an inherent property of the model. Motivated by the pursuit of positional-encoding-free architectures, this work explores a language model architecture that integrates causal state-space equations to implicitly encode positional information before attention computation. Specifically, each model block applies a causal state-space equation before self-attention, allowing recurrent state dynamics to encode sequential information into token representations. Consequently, subsequent attention layers operate on position-aware representations without requiring explicit positional encodings while retaining the expressive modeling capacity of self-attention. We present \textsc{ZetaGPT}, a compact hybrid language model designed for research, rapid prototyping, algorithm verification, and educational applications. In addition to the proposed architecture, \textsc{ZetaGPT} provides a fully open-source, end-to-end training pipeline encompassing dataset construction, tokenizer training, pretraining, supervised fine-tuning, reinforcement learning from human feedback (RLHF), and chain-of-thought (CoT) reasoning via pure reinforcement learning. To the best of our knowledge, \textsc{ZetaGPT} is the first open-source small language model without explicit positional encoding and establishes a compact, reproducible reference implementation for the development and empirical study of positional-encoding-free language models.

Autoregressive language models align training and use: generation conditions on a clean prompt, and training predicts future tokens from clean left context. Diffusion language models offer parallel denoising, but native dLLM pretraining can randomly corrupt prompt and continuation tokens together, weakening the clean-prefix interface needed for prompt-conditioned generation. We identify this mismatch for prompt continuation and propose PCD (Prefix-Conditioned Diffusion), a pretraining objective that combines AR prefix supervision with no-shift suffix denoising. At the training-objective level, PCD changes the attention mask, corruption mask, and label construction in continued pretraining; it does not require an autoregressive decoder, verifier, or new inference mode. By supervising the clean-prefix side autoregressively and applying diffusion only to the unknown continuation, PCD makes the local training interface resemble how block-diffusion models are queried at evaluation time. We further separate intra-sample prefix conditioning from inter-sample objective mixing, allowing us to identify the local alignment signal separately from the optional batch-level mixing knob. Across LLaDA2-Mini and Qwen-1.7B backbones, PCD consistently improves over same-family native dLLM stable baselines, reaching a 4.2% relative gain on the main LLaDA2-Mini six-benchmark average (+2.56 points) and a 14.2% relative gain in the primary Qwen mechanism comparison (+4.86 points). These results suggest that aligning the pretraining context distribution with prompt-conditioned generation can recover a measurable part of the dLLM continuation gap without changing inference.

Wearable human activity recognition (HAR) remains challenging due to the computational and energy constraints of deep learning models on resource-limited devices. Existing lightweight approaches often rely on recurrent architectures (e.g., GRU and LSTM), limiting parallelism and increasing inference latency. We propose LITEWAY, a modality-agnostic, fully convolutional framework for multichannel sensor time series that replaces recurrent temporal modeling with structured convolutional decomposition. LITEWAY combines lightweight convolutional blocks, strided temporal processing, and convolution-attention pooling to efficiently capture temporal dependencies while reducing computational complexity. We evaluate LITEWAY on 16 HAR datasets against TinyHAR, TinierHAR, and MLP-HAR. LITEWAY achieves competitive macro F1 while reducing model size by 4.06x-9.52x (Light) and 3.87x-9.07x (Full) compared with TinyHAR and TinierHAR. Deployment experiments further show energy reductions of 2.29x-3.14x (Light) and 1.46x-2.01x (Full) compared with TinierHAR and MLP-HAR, highlighting efficient fully convolutional temporal modeling for wearable HAR. The source code is publicly available at https://github.com/dominique-nshimyimana/liteway.

User simulators are widely used as scalable environments for training and evaluating interactive assistants. Generating the next user turn is inherently one-to-many: the same profile and dialogue context may support multiple plausible continuations with different local interaction intents. A fluent response may therefore advance the dialogue through an inappropriate intent, such as acceptance rather than repair. Our key insight is that controllable user simulation should separate which local interaction intent the next user turn should realize from how that intent is expressed in language. We introduce UserIDA (User Intent-Directive Alignment), which exposes interaction intent as an explicit per-turn directive. UserIDA defines a six-way intent interface, learns directive-conditioned generation through supervised fine-tuning, and uses intent-calibrated policy optimization during group-based reinforcement learning. The reward preserves composite response quality while ensuring that intent-violating candidates rank below compliant alternatives in mixed groups. On LMSYS-USP, UserIDA achieves 86.6\% intent accuracy, outperforming the strongest dedicated user-simulator baseline by 24.3 percentage points while improving semantic and stylistic similarity. In within-context interventions, it realizes at least four of the six target intents in 91.7\% of evaluated dialogue states, compared with 22.9\% for the strongest external baseline. These results establish per-turn intent control as a complementary dimension to response fidelity in user simulation.

Deep decoder-only Transformers often replace the original Post-Norm architecture with Pre-Norm variants because Post-Norm training is highly sensitive to warmup and learning rate under conventional initialization schemes. Although prior work has identified rank collapse and gradient vanishing as related symptoms, it remains poorly understood how causal attention creates high-similarity representations and why training dynamics fail to repair them. We give a two-stage analysis of Post-Norm rank collapse using token similarity as a scalar state variable. First, at initialization, causal attention acts approximately as a prefix-averaging operator that increases token similarity across depth, while the SwiGLU branch contributes only a smaller damping effect. Second, once training enters a high-similarity regime, growth of pre-normalization residual norms makes the RMSNorm backward factor contractive; under mild conditions, gradients to earlier layers decay geometrically. As a complementary result, we characterize the properties of a collapsed network: its best predictor is frequency distribution with relatively high loss floor, and gradients in collapsed layers vanish at frequency distribution. Experiments on 48-layer decoder-only Transformers trained on C4 dataset match the predicted initialization-time similarity growth and collapse-time gradient contraction, and show that collapsed runs stay near the predicted frequency loss. Together, these results distinguish the forward similarity amplification and backward repair incapacity in Post-Norm collapse, while also characterizing the behavior of collapsed networks.

KV-cache compression reduces long-context memory, but aggregate task scores reveal neither which correct executions fail nor why. We present KVDiagnosis, a diagnostic dataset and benchmark with three contributions. First, a 25-method taxonomy groups methods into five mechanism families and links them to eight verified implementations and their valid diagnostic measurements. Second, for every supported method setting, we evaluate all sources in each fixed split against a per-source FullCache control before selecting FullCache-correct/compressed-wrong (C-to-W) rows separately for each method-setting, so no compressor defines another's test set. Third, a common record format links paired outputs and run metadata to cache, likelihood, attention, and decoding measurements with explicit applicability states. On Qwen3-8B, four evidence-aware workloads yield 59 800 supported compressed runs over 2600 sources and 12 520 C-to-W rows. Under fixed diagnostic rules, 63.2% have low or partial measured/projected coverage. Only 19 rows (0.2%) combine high measured/projected coverage with strong likelihood drift; another 2,126 (17.0%) preserve structural position addressability, for which representation fidelity remains unknown, while showing the same drift. Against C-to-C success controls, all ten diagnostics separate failed from successful compression (stratified AUROC 0.684-0.871). Among 96 reproducible low-EAR failures, a controlled 4x evidence-attention boost repairs 29.2%, versus 6.3% under a count-matched sham intervention and 3.3% degradation on matched C-to-C controls. Code and data are available at https://github.com/ChosenQC/KVDiagnosis.

Research regarding the sign language recognition mostly relies on RGB images, whileas sign language datasets that provide depth images are limited. Point clouds obtained from depth images can be used for sign language recognition with neural networks like PointNet. In recent years, various neural networks are used for generating realistic depth images from monocular RGB images. In this work, synthetic depth images were created from RGB images using Depth Anything V2 network. For this purpose, three sign language datasets (Real-time ASL Fingerspelling, KArSL, AUTSL) which contain both RGB and depth images were used. Classification accuracies of the point cloud data created from both original and synthetic depth images using various PointNet architectures were measured for sign language recognition. From the original and synthetic point clouds, frame based, Point Gesture Map and Long Short Term Memory data models were used for classification and their performances were compared. In the results, both original and synthetic based data achieved acceptable performance in most models. In general, original depth based point cloud models performed better than synthetic ones, however in some models synthetic depth based models performed better than the originals.

Process discovery is one of the central challenges in process mining. Petri nets are particularly attractive because simple local constructs can express complex behavior, including concurrency. While their global behavior may be difficult to analyze, individual places can be efficiently characterized using monotonic properties, enabling bottom-up discovery. Unlike top-down approaches such as the Inductive Miner, which rely on predefined constructs for sequences, choices, loops, and concurrency, our approach allows such structures to emerge organically and can exploit the full expressive power of Petri nets, including free-choice constructs and long-term dependencies. The main challenge is the exponential number of candidate places and their combinations. We present the SPECpp framework which implements strategies to obtain high-quality models under time and resource constraints. SPECpp supports rapid experimentation and is used to evaluate these strategies using both synthetic and real-life event data.

Invariant learning seeks representations that remain predictive across environments, yet the behavior of its objectives along the regularization path is often opaque. We address this objective-behavior gap by viewing representation learning as multimode magnetization and deriving, from concrete invariant-learning objectives, a Landau-type effective free energy whose low-order coefficients form objective signatures and induce distinct regularization phenotypes. Effective quadratic corrections move the phase boundary and enable finite-strength mode elimination; quartic corrections regulate post-onset amplitude and typically leave residual loading at finite strength; higher-order structure governs non-monotone tails, instability, and collapse at large regularization. In a canonical bilinear model, the theory yields closed-form phase boundaries and steady-state loadings, as well as distinct critical strengths for shortcut and stable modes that define a selective-retention window. Controlled experiments confirm the predicted phase boundaries, loadings, and regularization phenotypes. In one- and two-hidden-layer ReLU networks, the same signatures remain predictive of qualitative regularization-path behavior despite depth-dependent shifts in scale. A matrix extension generalizes the framework to coupled collective modes and yields a spectral phase-boundary criterion. Together, the framework turns low-order objective signatures into predictions of regularization phenotypes and, ultimately, of what models learn as regularization varies.

We identify and quantify temporal misgrounding: the systematic retrieval and citation of the currently in-force version of a legal article when the applicable version is an earlier or future one. Standard legal RAG treats the corpus as static; we argue legal question answering is a temporally-indexed retrieval problem. We introduce FiscalQA Pro, pairing a versioned corpus of 32,436 article-versions of the French tax code (93 years, 1938-2031) with an all-model-hard temporal-reasoning track: 209 scored, expert-reviewed questions across 33 CGI articles (221 released; twelve flagged out of the answerable scope). At selection time, no evaluated model recovered its date-applicable answer closed-book in any of four sampling draws, and the currently in-force text lacks the gold value for all but one of the scored questions. Answers are scored deterministically via atomic ground-truth "nuggets" (regex and numeric-with-tolerance), never LLM-as-judge: an LLM judge would inherit the temporal bias it is meant to score. Across eleven models (five frontier closed-API systems plus Gemini 2.5 Pro as a substitute entry, and five open-weight), parametric knowledge yields 3.0% mean strict accuracy and RAG over a static current-version corpus 2.7%. Static RAG retrieves the date-applicable version 0% of the time, confidently citing a real but inapplicable version. Our end-to-end retriever over a multi-version index, with no oracle, reaches 98.3% mean strict; an oracle-article ablation reaches 99.1%, locating the residual gap in first-stage recall, not version selection. We additionally release a version-aware jurisprudence dataset of 69,208 citation links, together with the corpus, benchmark, model responses, and pipeline code.

This note aims to serve as an entry point to the literature on learning in games, a topic with significant theoretical appeal and a wide range of applications -- from machine learning and data science to economics and beyond. Our presentation is structured around two complementary viewpoints: We first consider a single agent -- the learner -- engaged in a sequential decision process in an unknown, non-stationary, and possibly adversarial environment. We then examine what happens when the environment is shaped by the decisions of several interacting agents, not necessarily aware of each other's actions or goals, and all seeking to improve their individual rewards. In this general context, we examine a family of regularized learning policies based on best-responding to the past history of play, up to a regularization penalty intended to encourage exploration and prevent over-commitment to suboptimal choices. In the single-agent setting, we present some basic regret bounds for regularized learning in adversarial multi-armed bandits; in the multi-agent setting, we describe an ergodic equilibrium convergence result for zero-sum games in the spirit of classical results on fictitious play, as well as a "folk theorem" linking strategic and dynamic notions of stability -- Nash equilibria and attracting points of regularized learning, respectively. We pay special attention to the information available to the players and, through a unified analysis framework, we study both oracle- and payoff-based (bandit) methods. Our goal is to provide a coherent and comprehensible -- albeit, by necessity, not comprehensive -- account of some recent ideas in the field, and to discuss their implications for the study of rationality.

Generative AI models are primarily designed to imitate the data distribution, an objective that neither corrects diversity lost by a learned generator nor defines how generation should extend beyond the diversity of the data itself. We introduce Imaginative Generative AI (IGA), a framework that makes diversity part of the target-distribution design problem: among distributions close to a reference, IGA selects one whose spectral diversity reaches a prescribed level. Diversity is measured by the von Neumann entropy of the generated distribution's kernel covariance operator in a fixed representation space, providing a reference-free representation-guided measure of how broadly probability mass occupies embedding directions. The spectral entropy of the population data distribution defines an Entropy Wall. Below the wall, IGA performs diversity repair, recovering variation that a learned generator has lost while remaining within the diversity level of the data. Beyond the wall, the data distribution itself becomes infeasible, and IGA deliberately departs from it to produce distributions with greater representation-relative spectral diversity, an operational notion of imaginative generation. These regimes form a single regularization path from imitation to imagination and define an i.i.d. target distribution at each prescribed diversity level. We develop the theory of this entropy-constrained projection and show that, under a KL anchor to a pretrained generator, the optimum satisfies a self-consistent exponential-tilt relation. This characterization leads to IGA Guidance, a retraining-free inference-time method for score-based and diffusion models, including DDPM and DDIM samplers. Experiments on synthetic and vision benchmarks demonstrate diversity repair below the Entropy Wall and controlled spectral extrapolation beyond it.

Electromagnetic inverse scattering is a nonlinear and ill-posed problem, where accurate reconstruction is challenging due to measurement limitations, noise, and high computational costs, especially for 3-D imaging. Although physics-driven neural networks (PDNNs) reduce the dependence on labeled training data, existing accelerated PDNN frameworks often rely on preliminary reconstruction-based region selection, which may introduce instability when the selected region is inaccurate. In this paper, a coordinate-residual physics-driven neural network (CRPDNN) is proposed for 3-D electromagnetic inverse scattering. The proposed solver directly reconstructs the unknown contrast distribution using normalized spatial coordinates and a residual convolutional network, without requiring a preliminary reconstruction. For the reported noise-free 3-D synthetic cases, CRPDNN achieves an average relative error of 2.10\%, compared with 7.97\% for CSI and 3.99\% for $L_{2/3}$-FBE-WCIE, while providing approximately 5.5- and 12.1-fold speedups over the two baselines, respectively. Supplementary 2-D comparisons further confirm its stability and computational efficiency relative to existing PDNN frameworks. CRPDNN also maintains reliable reconstruction performance under noisy measurements, and the 3-D Fresnel experiments further indicate its potential for practical imaging applications. The related code is available at https://github.com/Physics-driven-methods.

Long-horizon complex tasks require agents to repeatedly observe states, formulate plans, invoke tools, verify results, and recover from failures in continuously changing environments. However, such control experience often remains confined to a single context or a fixed prompt, and is difficult to accumulate and reuse across historical traces. This paper presents OpenLoopEvolve (OLE), a self-evolution framework centered on the Loop Policy. OLE represents an agent's observation, planning, memory, action, verification, recovery, stopping, and budget control as portable policy assets with versions and lineages, and provides online and offline evolution modes that can be selected according to practical needs: the online mode triggers candidate generation based on feedback from continuous operation, whereas the offline mode searches for candidate policies from archived traces and failure evidence. Both modes share an evolution mechanism consisting of autonomous proposals by a large language model, Champion--Challenger paired evaluation, and robust release. Policies released online are activated at a subsequent task boundary, monitored using subsequent feedback, and rolled back to their parent versions when degradation conditions are met. On the simulated business benchmark YC-Bench, both modes improve aggregate task performance, task success rate, and risk metrics relative to a fixed initial Loop Policy. The results indicate that treating the Loop Policy as a governable asset can support the accumulation, comparison, release, and reuse of control experience and improve agent performance on long-horizon complex tasks.

Electrical circuit analysis requires more than recognizing components in an image. A solver must ground symbols and labels, recover latent topology, select a physical model, formulate coupled equations, propagate intermediate quantities, and preserve units, signs, directions, and phase conventions. We introduce \benchmark, a benchmark of 1,000 authentic textbook problems for evaluating this complete long-horizon visual-to-symbolic reasoning process. Each problem pairs one or more circuit diagrams with a self-contained question, a typed or semantically specified answer, and a reference worked solution. An evidence-first construction pipeline aligns questions, figures, and solutions, while a reasoning-oriented taxonomy organizes problems by circuit type and dependency depth. Evaluation combines conservative typed scoring with identity-blinded multi-model semantic consensus, retaining every problem in the denominator. Across three commercial chatbot systems and six open-source multimodal large language models, the highest-scoring system reaches 84.8\% accuracy. However, performance consistently deteriorates on long-horizon problems, and qualitative analysis exposes persistent failures in topology-to-target binding, physical conventions, and late-stage output propagation. \benchmark{} provides a focused testbed for measuring whether multimodal models can transform technical visual evidence into sustained, physically valid symbolic reasoning. Code are available at GitHub - CircuitReason/CircuitReason1K.

Visual object tracking requires effective temporal integration, yet most Transformer trackers still rely on predominantly feed-forward feature extraction. Existing temporal mechanisms typically update templates, prompts, queries, or prediction states, while intermediate representations are rarely reused to modulate corresponding processing stages. We propose \textbf{FeedbackTrack}, a visual-cortex-inspired framework that introduces sparse, group-level layer-aligned cross-frame feedback into pretrained Transformer trackers. Previous-frame intermediate states are detached, cached, and returned to corresponding Transformer groups in the current frame through two lightweight pathways: Query Feedback for token-level query modulation and Gate Feedback for context-dependent feature modulation. FeedbackTrack preserves the original tracking pipeline with only a fixed-size one-frame cache. Across SPMTrack and ARTrackV2, FeedbackTrack consistently improves five backbone configurations on LaSOT and GOT-10k, achieving 83.4 AO and 79.1 AUC with SPMTrack-G while adding less than 1\% parameters. Controlled comparisons show that cross-frame feedback outperforms same-frame modulation by 1.8--3.2 AO points, demonstrating that the gains mainly come from recurrent historical information. Further analysis reveals a non-uniform depth-dependent organization of learned feedback strengths, highlighting the effectiveness of recurrent feedback for Transformer tracking.

Large-scale learning systems often face the challenge of balancing multiple, potentially competing objectives, such as fairness, accuracy, and latency. While recent work has formalized this as an optimization problem over binary states, many real-world control parameters, such as fairness thresholds, diversity mixing rates, or resource budgets, are continuous. In this work, we extend the framework to \emph{continuous state spaces}. We model the problem as minimizing a sum of linear objectives subject to \emph{movement costs} that penalize system instability. We capture the local structure of the objectives using a \emph{dependency graph} (or factor graph), where each objective is determined by a subset of the state attributes. To address the tension between exploration and stability, we propose \emph{Lazy Graph-LinUCB}, an algorithm that performs lazy updates to minimize switching costs while maintaining near-optimal regret. Beyond stability, we introduce three advanced mechanisms to exploit the underlying graph structure: (1) an \emph{asynchronous} update schedule that eliminates synchronization overhead in sparse graphs; (2) an \emph{adaptive} algorithm that learns the graph structure from data; and (3) a \emph{joint estimator} that leverages data sharing among correlated objectives to significantly tighten regret bounds. Empirically, we demonstrate that these structural exploitations reduce movement costs by more than a factor of three in heterogeneous systems while maintaining similar cumulative losses.

The demand for maritime surveillance has given rise to the need for monitoring fishing vessel activities, particularly in addressing the challenge of "dark vessels" that operate without Automatic Identification System (AIS) transmission. This study presents a novel approach for detecting small-scale fishing vessels using nighttime light (NTL) imagery from the SDGSAT-1 satellite, combined with deep learning techniques to enhance fishing monitoring awareness along the western coast of India. A dual-branch YOLO11 architecture was developed to exploit both the 10-meter panchromatic and 40-meter RGB imagery from SDGSAT-1. The custom model architecture was specifically optimized for small object detection in NTL imagery, featuring parallel convolutional backbones that process both modalities before concatenation for enhanced feature extraction. The dual-branch YOLO11 model demonstrated optimal performance with a precision of 0.99, recall of 0.93, F1-score of 0.96, and mAP@50 of 0.96, significantly outperforming single-branch implementations of YOLOv5s, YOLOv8s, and standard YOLO11s architectures. When applied to the western coast of India, the model detected 31525 vessel instances across the temporal dataset spanning 2022-23. Cross-matching analysis with AIS data revealed that only 7146 (22.7%) of detected vessels had corresponding AIS transmissions, while 24379 (77.3%) were identified as potential dark vessels. Spatio-temporal analysis showed peak fishing activity during January-April, with a primary activity corridor parallel to the coastline within 50-100 km, corresponding to productive continental shelf areas. This research contributes to maritime surveillance capabilities by highlighting the effectiveness of nighttime lights satellite imagery for fishing vessel detection and provides valuable insights into fishing patterns and potential regulatory compliance issues in Indian waters.

Closely related languages written in different scripts expose little surface overlap to multilingual models, limiting cross-lingual transfer. We compare two approaches to script unification: the general-purpose uroman romanizer and the family-specific Common Turkic Script (CTS). We train matched fastText models on transliterated Wikipedia corpora from 11 Turkic languages and evaluate them on WikiANN named entity recognition and Universal Dependencies part-of-speech tagging. CTS and uroman show no significant difference on NER, while both substantially outperform the official monolingual fastText baselines. POS results reveal no universal winner: language-specific differences are associated with the cross-lingual character n-gram coverage induced by each representation, while within-language coverage becomes more important when target-language supervision is available. Although CANINE-c achieves higher overall POS averages, the substantially simpler fastText-based systems remain competitive on several treebanks. Overall, the effectiveness of script unification depends on the language, the induced subword overlap, and the available supervision.

Test-time scaling improves LLM accuracy but multiplies inference cost, making the accuracy gained per unit of compute the metric that matters in deployment. Self-consistency is one of the established approaches, which spends this budget entirely on the output side by sampling repeated reasoning paths. We study Test-Time Augmentation (TTA), which extends self-consistency by also perturbing the input, aggregating predictions across transformed versions of the input, and ask whether input-side diversity converts compute into accuracy more efficiently than output-side diversity. We perform a systematic, matched-compute comparison: we evaluate three simple input-side strategies (semantic rephrasing, lexical perturbations, and visual transformations) across six datasets covering general and multilingual knowledge, mathematical reasoning, multi-modal question answering, and sentiment classification, against chain-of-thought prompting and self-consistency. Semantic rephrasing delivers consistent and statistically significant accuracy gains while Pareto-dominating self-consistency on cost-effectiveness, delivering roughly 1.8X more accuracy per dollar and outperforming it on five of six tasks. We further analyze the number of augmentations, multi-modal strategies, and base model scaling, finding that TTA is most cost-effective for mid-tier models where a stronger model is unavailable or too expensive. Our findings indicate that for current mid-tier LLMs, varying the input converts inference compute into accuracy more efficiently than varying the reasoning path alone. The TTA implementation is available at https://github.com/aws-samples/sample-genai-reflection-for-bedrock.

The Lunar Reconnaissance Orbiter (LRO) has been collecting high-resolution images (at around 0.5-2 meters per pixel linearly with its Narrow Angle Camera) of the Moon since 2009, amassing a large dataset of images and offering researchers the opportunity to study the surface of the Moon at unprecedented scale. Here, we aim to test the abilities of the Beta-Variational Autoencoder (VAE) created by Lesnikowski et al. (2024), an unsupervised learning model which identifies anomalous features across the Moon's surface, locating not only scientifically useful geologic formations such as rockfall deposits, fresh impact craters, irregular mare patches, or volcanic pits/collapsed lava tubes, but also artificial objects such as landed spacecraft. This investigation further gauged the model's ability to locate anomalous surface features, successfully recovering two places of interest (Plaskett Crater and Paracelsus C Crater) and numerous landed technological assets at a statistically significant rate.

Military capability is increasingly determined by software. Yet defence platforms are procured on decade-long timescales, while the software and AI models they carry must evolve in days or hours. This paper calls this mismatch the lifecycle paradox, and argues it is the central problem Software-Defined Defence (SDD) must solve. SDD rests on three dimensions: software and systems engineering (design, procurement, certification), AI engineering (sovereignty and trust of learned components), and connectivity and infrastructure engineering (timely exchange of information among sensors, AI, and operators). The proposed path to resilient SDD starts from civilian technologies, addressed through a continuous, DevOps-style loop: model-based systems engineering and simulation-based testing front-load design and verification; tactical connectivity and low-power edge execution carry that design into contested operation; continuous compliance, assurance, and variability management run as cross-cutting concerns. This loop is sustainable given capabilities already proven in automotive, manufacturing, space, and energy. The next step is validating them under adversarial or defence-certified conditions, with short-, medium-, and long-term paths to closing gaps. Closing the SDD gap while preserving civic benefits is a distributed responsibility: researchers must redirect methods toward adversarial conditions; industry must expose tooling to operational needs; policymakers must shape regulatory instruments; and defence agencies must validate results with operators. Recommendations span three horizons: a short-term baseline of adversarial testing and connectivity pilots; a medium-term pipeline of incremental certification; and a long-term validation closing the loop under operational conditions.

Density estimation underlies many unsupervised tasks on tabular data such as anomaly detection, out-of-distribution detection, and data augmentation. Although all these problems reduce to questions about where probability mass lies, they are typically solved individually by fitting a separate model to each dataset, with its own hyperparameters and tuning budget. We introduce ICED, an in-context, energy-based density estimator that removes this per-dataset cost. ICED is a transformer-based model pretrained once on a synthetic prior built specifically for density estimation under an objective that fits log-density where it is informative and preserves its ordering elsewhere. In the inference, it reads a dataset as context and returns an unnormalized log-density for any query point in a single forward pass, with no fitting, sampling, or hyperparameter selection. A single frozen ICED model then drives four tasks usually handled by four specialized pipelines: density estimation, out-of-distribution detection, unsupervised anomaly detection, and generative augmentation. Across all four, it is competitive with the strongest task-specific method, while being the only approach that needs no retraining, no tuning, and no labels to move between them. The code is available at https://github.com/gmum/iced.

Simulation-based optimization (SBO) evaluates executable policies under stochastic dynamics, but most methods treat the simulator as a black box: aggregate scores rank candidates without revealing why they fail or which policy logic should change. We present an LLM-guided heuristic design framework that uses repeated simulation for selection and event-level traces for diagnosis. Each incumbent is assessed through multiple replications, while replaying its lowest-scoring one produces a queryable trace. A manager agent formulates bottleneck hypotheses from this evidence, and editing agents implement parallel code-level revisions. After execution checks and repeated evaluation, best-so-far selection retains only improvements. LLM revision occurs between evaluation batches, while a fixed policy controls each simulation run. We evaluate the framework in a discrete-event simulation of dynamic production and automated guided vehicle (AGV) scheduling. Across five independent optimization runs with Gemini-3.1-Pro, final mean scores averaged 77.51 on the simulator's 0-100 scale. In the highest-scoring run, trace-based diagnoses motivated proactive charging, distance-aware AGV assignment, and rebalanced dispatch priorities, raising the best-so-far mean score from 62.49 to 78.61. On 100 matched seeds, the best final policy outscored representative rolling-MILP, rule-based, and metaheuristic policies on every seed and retained its advantage under random faults without re-optimization. After separate re-optimization for a longer horizon and variable order interarrival times, the resulting policies again outscored all baselines. Ablations with two LLM backbones showed that removing either parallel candidate generation or trace-database access reduced final mean scores. These results show that simulation traces can guide targeted code-level policy improvement in complex simulation-based scheduling.

Multistage stochastic model predictive control (MPC) handles uncertainty by optimizing over a scenario tree, a finite branching approximation of future outcomes constructed from sampled forecasts. To build such a tree, conventional methods focus on matching the underlying probability distribution---e.g., via Wasserstein-based scenario reduction---but improved distributional accuracy does not necessarily yield better control performance. We propose a control-oriented approach that learns scenario tree construction directly from its impact on downstream decisions. Fixing the tree topology, we formulate tree construction as a sequential assignment of sampled scenarios to leaves. This assignment is parameterized by an attention-based policy over the scenario set and trained using reinforcement learning, with closed-loop control profit as the objective. Training is stabilized by an asymmetric critic that leverages realized future trajectories. We evaluate the method on a risk-averse battery arbitrage problem. Across a range of forecast set sizes, the learned construction consistently achieves the highest profit, outperforming classical forward and backward reduction methods and certainty-equivalent (single-trajectory forecast) control. The learned policy also exhibits greater robustness on challenging instances, consistently demonstrating better tail-risk characteristics. Analysis of the resulting trees indicates that our method constructs compact, selectively branching structures that capture high-impact events while keeping most trajectories nearly deterministic. These findings highlight that the value of a scenario tree depends critically on the decisions it supports, and provide an effective framework to train scenario tree constructors merely based on the closed-loop control optimization signal.

Hallucinations are a major concern for the integration of artificial intelligence into medicine, although less explored in the realm of medical image processing. Unlike problems in natural text understanding and reasoning therewith, determining whether or not predictions derived from biomedical images and signals is less intuitively clear. This article suggests that topological errors could constitute hallucinations in a way that can be more readily measured and thus regulated. Certain of these properties for certain types of problems, such as biomedical signal segmentation, can be rephrased as linear temporal logic predicates, a number of which can be explicitly enforced using probabilistic graphical models. Our simulations show the potential of these explicitly constrained predicates for the case of automatic surgical phase recognition in robot-assisted hysterectomy, improving accuracy by approximately 10% while removing the vast majority of topological errors, suggesting that mathematical guarantees of correctness can supplement other empirical forms of regulating machine learning in medical image computing and computer-assisted interventions.

Brain-to-audio reconstruction is limited by \emph{prior domination}: when a pretrained generator is conditioned on a weak neural signal, it produces realistic but stimulus-inaccurate audio. We introduce RAG-Audio, which decodes fMRI into a semantic audio embedding, retrieves a matching real-audio exemplar, and initializes the frozen generator's sampling trajectory from that exemplar while retaining the decoded embedding as conditioning. On Brain2Music, RAG-Audio improves 10-way stimulus identification from $0.14$--$0.18$ for direct generation, near the $0.10$ chance level, to $0.40$--$0.43$, comparable to retrieval. It also reduces Fréchet Audio Distance by roughly an order of magnitude, from $13.49$ to $1.25$ for AudioLDM. RAG-Audio approaches nearest-neighbor retrieval in identification while remaining generative; its higher FAD is expected because retrieval directly replays real audio. An autoregressive negative control, which lacks an initializable latent trajectory, shows no comparable gain, attributing the improvement to trajectory initialization. These results suggest that retrieval-guided initialization can mitigate prior domination in brain-to-audio generation.

Model learning by an eavesdropper is treated as an estimation problem in a federated environment. The Fisher Information Matrix for the eavesdropper's estimation problem is driven to singularity through a signaling design; this ensures that the eavesdropper cannot learn the model. Herein, the innovation of prior designs is that model shifts are designed to maximize the difference in the model learned by Eve and the central server while satisfying a transmission power constraint for the agents. Two shift schemes are provided. MaxModShift outperforms a prior ModShift design while requiring lesser transmission power. Compared to a noise injection scheme, MaxModShift performs better while requiring a lower bandwidth secret channel and a reduced average power consumption.

Newly triggered landslides rarely carry immediate annotations, so cross-domain transferability determines the value of landslide mapping for emergency response and regional risk assessment. Vision foundation models have strengthened representational transfer, yet on unseen regions, events, and data sources they still generate high-confidence false alarms. Terrain, material, and rainfall triggering can constrain such errors, but their supports are local, regional, and event-scale, so that resampling onto a 10~m grid misaligns them with the segmentation decision unit and compounds the uncertain geographic context problem (UGCoP). We propose GeoPhysAdapter, which anchors on a frozen vision foundation model, restricts terrain, material, and triggering to dense spatial guidance, regional modulation, and event-timing forcing, and applies bounded adaptation at two decision units, the pixel and the candidate landslide body, reverting exactly to the visual prediction where support is insufficient. On an event-isolated PILD dataset of four public sources, 55 global landslide events, and 7,890 test samples, 70.3% of cross-domain false-positive mass lies in near-pure spurious bodies of median equivalent diameter 207m, matching coarse-prior support rather than the pixel. Pixel-level adaptation removes a net 507,817 erroneous pixels and reduces error by 7.76%, whereas raising the decision unit to the candidate body, under identical samples, anchor, and baseline, increases error reduction to 23.99%, approximately 3.1 times the pixel-level effect, improves IoU by 0.031 (14.2% relative), and corrects 9.92 pixels per pixel harmed. The data and code are publicly available at: https://github.com/Liu-Zhihang/geophysadapter.

On unlabeled test data, reinforcement learning lacks a ground-truth reward; test-time RL methods derive one from the model's own roll-outs, rewarding those that match the majority vote over $N$ sampled answers. That vote discards a correct answer whenever it is a minority and scores every majority-matching roll-out identically. We replace it with \emph{CoRE} (Consensus Rewards via Equilibrium): the $N$ roll-outs form a graph whose edges combine answer agreement, reasoning similarity, and generation confidence, and replicator dynamics extract its dominant set, yielding a refined pseudo-label, a graded per-roll-out reward, and a per-question cohesiveness gate. CoRE strictly generalizes voting: majority voting is recovered as a special case; a block-value analysis gives a sharp threshold for when consensus recovers a correct minority against a larger wrong plurality; and confidence calibration provably lowers that threshold multiplicatively. Across seven backbones and five benchmarks (42 model--benchmark cells, three seeds each), \emph{CoRE} improves the untrained base by $+21.7$ points on average versus $+20.4$ for majority-vote TTRL, wins wherever agreement is contestable with margins over the vote of up to $+7.5$ points, and reaches the voting baseline's plateau accuracy in $54$--$70$\% fewer steps. Consensus, not counting: treating the roll-out group as a graph rather than a ballot box turns a brittle vote into a calibrated, graded, self-supervised reward at no extra roll-out cost.

Quantum mutation analysis is emerging as an essential technique for evaluating test suites due to the limited availability of real faulty quantum programs. However, existing quantum mutation analysis tools use fixed gate-based mutations, resulting in easy-to-detect mutants, which reduces their effectiveness in assessing the quality of test suites. We propose QUMUG, a search-based approach for generating challenging mutants by utilising parameterisable quantum gates. QUMUG employs search algorithms to optimise mutation parameters and find non-equivalent mutants passing a given test suite. In our evaluation over 30 quantum programs, QUMUG produced mutants that are three times more challenging than the mutants generated by existing tools. Among the four evaluated search algorithms, the genetic algorithm was the most effective, generating an average of 494 undetected mutants per program with a 99.67% success rate and 94.3% non-equivalent ratio. The generated mutants demonstrated their effectiveness by requiring the addition of five times more test cases to the test suite than the mutants generated by existing tools. We also analysed the behaviour of higher order mutants in quantum circuits, and showed that while first order mutations are more effective for enhancing the test suite, higher order mutants highlight the need for new unique test cases.

While mathematical models act as vital decision support systems for operational Air Traffic Flow and Capacity Management (ATFCM), existing approaches isolate Air Traffic Flow Management (ATFM) from Dynamic Airspace Configuration (DAC). This separation introduces an unresolved circular dependency between fixed-demand and fixed-capacity assumptions. Although joint optimization resolves this gap, the enlarged search space renders exact models computationally intractable for medium- to large-scale instances. To bridge this gap, we propose ASPaeroFlow: a heuristic for the joint ATFCM; it combines instance-space decomposition heuristics with a local exact approach using Answer Set Programming. We benchmark ASPaeroFlow from small to industry-sized instances and compare it with exact and alternative approaches. The results indicate that (1) the heuristic provides a computational middle ground between exact methods and operational baselines; (2) simultaneous optimization can outperform sequential optimization on joint ATFCM; and (3) an ablation study indicates that DAC has a larger impact on solution quality than flow measures.

In a federated learning setup for GANs, several adversarial attacks are possible. One such attack is label flipping, in which malicious clients deliberately alter label information during local training in order to manipulate the global generator. The objective of this attack is to skew the learned generation distribution so that samples conditioned on a target label are instead mapped to a source class. In this work, we investigate the effectiveness of label flipping attacks in federated GANs through both theoretical analysis and empirical evaluation. We further consider an oversampling based variant, in which malicious clients upweight poisoned samples during local training to amplify their influence on the aggregated global model. We quantify the resulting distributional shift by computing the Kullback Leibler divergence between the clean and poisoned class conditional distributions, and show both analytically and on FEMNIST, MNIST, and CIFAR10 that the semantic damage of the attack grows linearly in the effective poisoning strength while deviation from the true target distribution grows only quadratically, making the attack effective yet difficult to detect from label agnostic metrics.

We present a linearized form of 2-simplicial attention by rewriting the trilinear score as an inner product between a composite query and a key, so that the sum over one token axis takes the same form as ordinary softmax attention. We then approximate this sum with positive random features and store the entire past in a fixed-size state, while the second axis stays explicit over a short window of recent tokens. This enables us to achieve linear cost in sequence length combined with a global reach that windowed 2-simplicial attention lacks. We implement it with custom Triton kernels and combine it with Kimi Delta Attention to build a model with no softmax attention at all. Under matched compute, this model achieves the highest mean downstream accuracy among the compared architectures, and at 16k context it improves mean accuracy over a KDA hybrid while lowering LAMBADA perplexity from 715.6 to 602.6.

The deployment of autonomous robotic systems in chemistry laboratories is accelerating experimental workflows and providing the foundational data for AI-driven scientific discovery. However, despite the success of data-driven methods in acquiring dexterous skills, safety remains a primary barrier to their deployment in high-risk domains, such as early-stage materials chemistry experiments. Specifically, learning-based policies frequently struggle to distinguish between safe and unsafe actions, leading to overconfident extrapolation and potentially catastrophic failures. To mitigate these safety risks, we propose SAFE-CHEM, an uncertainty-aware framework designed for robust, learning-based robotic chemists. Our approach leverages an ensemble of recurrent neural network-based imitation learning policies to quantify epistemic uncertainty online through the variance of action predictions. By characterising the success-conditioned density of this variance using kernel density estimation, we introduce a hybrid control architecture that autonomously switches from the learned policy to a deterministic, rule-based backup controller when uncertainty exceeds a calibrated safety threshold. We evaluate SAFE-CHEM across three fundamental laboratory manipulation tasks, where our empirical results demonstrate that this hybrid strategy improves overall task success rates and reduces critical safety violations compared to traditional single-policy baselines. Finally, we demonstrate the practical viability of the framework through zero-shot sim-to-real transfer onto a physical Franka Production 3 robot manipulator.

Action-conditioned world models (ACWMs) promise to provide embodied AI with scalable predictive simulators for planning, policy evaluation, and data generation. Realizing this promise requires precise action-conditioned transitions rather than merely plausible outputs. Yet their applicability remains difficult to establish because prevailing evaluations emphasize visual quality, task outcomes, or coarse rollout-level responsiveness without directly testing simulator fidelity. To address this gap, we evaluate ACWMs through the observable capabilities expected of physical simulators. Accordingly, we formalize Observable Simulator Contract, a minimal contract that any action-conditioned physical simulator should satisfy: supplied actions must induce corresponding agent motion, and environment responses must be grounded in that realized motion. To operationalize this contract, we introduce WorldSimProbe, comprising five controlled suites spanning local control sensitivity, global trajectory variation, source-diverse actions, interaction grounding, and dynamics. Suite-specific evaluators assess simulator-relative calibration, dense action-to-motion correspondence, false-interaction grounding, and primitive-level dynamics. We evaluate six open-source ACWMs on more than 18,000 instances across RoboTwin, ManiSkill, and LIBERO. World-SimProbe reveals systematic action-realization degradation across control variation, structured failures in interaction grounding and dynamics, and benchmark signals consistent with human judgments and downstream outcomes. Together, this capability-based framework provides a transparent, and standardized paradigm for diagnosing ACWM simulator fidelity beyond coarse, task-directed evaluation.

A CAD model is not engineering-grade merely because it looks correct. It must satisfy design requirements, respond predictably to parameter changes, support controlled edits, match a reference structural response under a declared analysis, and connect to other parts through valid joints. We present CADEngBench, a two-track benchmark for these capabilities. CADEngBench-P evaluates 300 parametric parts, each used for one zero-to-CAD task and one functional-editing task (600 tasks in total), through boundary-representation (B-Rep) validity, engineering and DFM checks, parameter-family perturbations, functional editing, and matched linear-static FEA in CalculiX. CADEngBench-A evaluates 150 body pairs through ranked joint retrieval, exact face-and-edge grounding, joint-frame prediction, and kinematic verification. Across eight multimodal, code-capable models, editing supplied CAD is substantially easier than generating it, while complex edits and matched FEA remain difficult. Assembly predictions often locate the relevant region but fail to recover the recorded joint or mating entities. These results show that CAD evaluation must test engineering behavior rather than appearance alone.

Self-evolving methods improve the capabilities of LLM agents by sampling trajectories from the underlying LLMs and learning from these trajectories. However, these methods struggle to learn beyond the inherent capability boundary of the agents, since the agents cannot sample correct trajectories on difficult examples for further improvements. In this paper, we propose a zeroth-order self-evolution framework that enables agents to learn beyond their capability boundary by perturbing LLM parameters to adapt to difficult examples without any trajectory annotations. Specifically, we perturb LoRA parameters of LLMs, run the agent, compute the losses under the perturbed and original parameters, and use the loss difference to estimate gradients and further update the LoRA parameters. We sample trajectories using the updated LLMs for supervised fine-tuning to break through the capability boundary of the agents, forming a closed self-evolution loop. We introduce a parallel perturbation inference mechanism and an adaptive lookup mechanism to reduce time consumption in zeroth-order optimization, with an answer perplexity loss that provides smooth and stable zeroth-order loss values. Experiments on multiple deep research benchmarks show that our method obtains substantially more successful trajectories and consistently outperforms strong baselines, especially on difficult examples. The code and released artifacts are available at https://github.com/hidk1911/ZOForLLMAgents.

LLM-based code review agents promise scalable, always-on review, yet current systems suffer from two intertwined weaknesses: (1) non-determinism--unbounded tool use makes review outcomes unstable, and (2) context locality--the reviewer's access remains bounded to the diff, capping discoverable issue depth. Both give rise to three challenges: misaligned context retrieval, a coherence-efficiency trade-off in multi-file pull requests, and hallucinated comments that erode trust. To address these, we introduce OpenCodeReview, built on deterministic engineering for uncertain agents: rather than granting maximal freedom, we inject determinism at three deliberate pipeline points. Rule-Guided Dispatch uses a multi-layer rule system to deterministically select files and review criteria, eliminating variability of agent-driven triage. Grounded File Review replaces free-form exploration with a curated tool set exposed through a ReAct loop, while file-level parallel SubAgents balance context coherence against efficiency and recover cross-file dependencies on demand. Independent Reflection introduces a falsification-first filter under an asymmetric information boundary--the reflector sees only the diff, not the agent's tool-augmented exploration--removing hallucinated comments without self-reinforcing bias, improving precision while preserving recall. On AACR-Bench (200 real-world PRs, 10 languages, 1,505 expert-verified comments), OpenCodeReview outperforms mainstream coding agents (e.g., Claude Code and Codex) across six LLM backends, achieving up to 2.17x higher SEM-F1 (25.10% vs. 11.57%) while consuming 5-15x fewer tokens. We open-source OpenCodeReview at https://github.com/alibaba/open-code-review.

Second language writing research distinguishes grammatical accuracy from native-like idiomaticity, yet automated writing evaluation often conflates these dimensions. This study introduces a layered LLM-correction pipeline that isolates structural errors from unnaturalness by generating literal error corrections and idiomatic revisions for 3,830 English writing samples from 120 Japanese junior high school students. Applying the regex-based CEFR-J grammar extractor, we quantify two diagnostic measures: accuracy gaps (structures attempted but incorrectly produced) and idiomatic gaps (grammatically correct structures underused or overused relative to native norms). Results reveal distinct patterns: definite articles, third-person singular -s, and modals (would, could) exhibit significant accuracy difficulties, while -ing forms and hypothetical modals (would) show the largest idiomatic underuse, with simple present verbs, subject-verb-object patterns, and modal can conversely exhibiting the most pronounced overuse. A two-dimensional instructional typology maps error rates against idiomatic gaps, distinguishing accurate but overused grammar items from error-prone or avoided complex forms requiring targeted production practice. This framework advances pedagogical feedback by enabling teachers to diagnose whether learner difficulties arise from inaccurate execution, structural avoidance, or L1-mapped overreliance, supporting evidence-based interventions tailored to the specific needs of each learner.

Audio-visual speech enhancement under real-world conditions remains challenging due to unreliable visual inputs and the lack of large-scale training data with realistic acoustic conditions. Existing approaches usually fuse visual features directly into the separation network, making them vulnerable to degraded visual signals. In this paper, we present DAVE, a decoupled audio-visual enhancement framework for real-world speech separation. Firstly, to address the data scarcity issue, we construct DAVE-Corpus, a large-scale training corpus with 219,411 mixtures generated from public meeting corpora through combinatorial acoustic augmentation. Then, we introduce a progressive multi-objective optimization strategy to jointly improve speech separation, intelligibility, speaker identity preservation, and perceptual quality. We further develop a certified selective enhancement chain that applies scene routing, GAN-based denoising, and loudness normalization only within the no-reference partition, guaranteeing non-degradation of reference-based metrics. Experimental results on the Real-World Audio-Visual Speech Enhancement Challenge demonstrate the robustness of DAVE under both real-world mixed scenarios and visual degradation conditions.

Data-Free Knowledge Distillation (DFKD) transfers knowledge from a pretrained teacher model to a compact student model by synthesizing semantically informative data, eliminating the need for access to the original training dataset. Existing DFKD methods rely heavily on architecture-specific statistical priors (e.g., Batch Normalization statistics) to guide data synthesis, however, such architecture-dependent priors are often absent in modern architectures such as Vision Transformers (ViTs), resulting in degraded semantic quality of the synthesized data and consequently catastrophic performance degradation. In this paper, we propose \emph{UniDFKD}, a unified data-free knowledge distillation framework that replaces architecture-specific statistics with explicit, architecture-agnostic semantic priors. \emph{UniDFKD} governs the entire synthesis-distillation pipeline along three dimensions: (1) Categorical Semantic Conditioning (CSC) defines \emph{what} to synthesize by persistently modulating the generator with language-derived embeddings to capture semantic diversity; (2) Spatial Semantic Anchoring (SSA) dictates \emph{where} evidence belongs by anchoring the teacher's spatial attributions to a Gaussian prior; and (3) Spatial Semantic Distillation (SSD) controls \emph{how} knowledge is transferred by explicitly aligning teacher-student spatial evidence alongside predictions. Extensive experiments across CNNs and ViTs demonstrate that UniDFKD establishes a new state-of-the-art, outperforming existing methods by an average absolute margin of over 20\% in both homogeneous and heterogeneous settings.

Global medium-range weather forecasting requires modeling structured yet state-dependent interactions among heterogeneous atmospheric fields. Existing data-driven models largely learn these interactions implicitly, whereas equation-level physical constraints may inherit approximation and model-form biases. We present VeinCast, a physics-guided dynamic field graph and graph-conditioned fusion framework that jointly forecasts 69 surface and upper-air fields. Within each local window, its Physics-Guided Dynamic Field Graph combines predefined atmospheric relations with state-dependent Top-K residual edges and adapts Earth-window attention using the resulting graph context. Graph-Conditioned Latent Fusion further employs graph context and source-node centrality to guide field-to-latent aggregation, while bounded feedback preserves field-specific information. On the $1.5^\circ$ ERA5 benchmark, VeinCast demonstrates competitive forecasting performance across all 69 meteorological fields at lead times of up to 14 days, compared with representative global weather forecasting models including FuXi, Pangu-Weather, GraphCast, FengWu, and ARROW. Ablations confirm that the two modules provide complementary gains, demonstrating the effectiveness of relational-level physical guidance for data-driven weather forecasting.

Learning-based wireless localizers often fail to utilize geometric information about the propagation environment, limiting their ability to exploit non-line-of-sight (NLoS) propagation and generalize across scenes. To bridge this gap, we propose GLocFM, a Geometry-aware Localization Foundation Model, which jointly exploits WiFi measurements and scene geometry represented as a 3D point cloud. We formulate localization as a maximum-likelihood (ML) estimation problem, where the goal is to find a transmitter position that maximizes the likelihood of the wireless observations conditioned on the scene geometry. The likelihood of a candidate transmitter position is calculated by a learned scoring function that matches the observed delay--angle-of-arrival (AoA) spectrum against the spectrum predicted for that candidate. A hierarchical scene encoder extracts propagation-relevant features to produce geometric priors for LoS and one-bounce reflection paths. For scenarios with imperfect synchronization, we further introduce a time-of-flight (ToF)-robust GLocFM model to handle unknown ToF offsets. GLocFM is trained on a multi-modal synthetic indoor localization dataset comprising 221 diverse scenes whose associated wireless signals are generated using Sionna RT. On both synthetic and the NeRF$^{2}$ dataset based on real measurements, GLocFM reduces mean 3D localization error relative to one of the state-of-the-art localization baselines by 49.5\% and 48.8\%, respectively. Ablations across different number of receiver, bandwidths, and array sizes further demonstrate the effectiveness and robustness of the proposed framework.

Real-world shopping often requires constructing a basket of complementary items rather than retrieving a single product. Such combo-shopping tasks arise in device setup, meal preparation, event planning, and group takeout ordering, requiring joint reasoning about item compatibility, availability, store-level requirements, delivery fees, coupons, and budgets. Evaluation is challenging because multiple baskets may satisfy the same request, making exact-match metrics unsuitable, whereas semantic evaluation alone cannot detect infeasible orders, invalid coupon combinations, or incorrect payments. We introduce ComboShoppingBench, an agentic shopping benchmark for open-ended yet verifiable basket construction in a simulated commerce and takeout environment. During task synthesis, an exploration agent constructs a feasible and semantically coherent basket of purchasable products; this witness guides the generation of coupons, budget constraints, user queries, and aligned evaluation rubrics. During evaluation, LLM judges assess semantic satisfaction, response quality, and claim faithfulness, while deterministic validation checks product-ID validity, budget compliance, and coupon optimality. Experiments with diverse LLM agents demonstrate that even strong agents struggle on ComboShoppingBench, highlighting substantial room for improvement in reliable, constraint-aware combo shopping.

Multimodal large language models (MLLMs) perform strongly on engineering imagery, yet existing benchmarks mostly test drawing recognition, information extraction, or compliance checking, leaving open whether models can combine distributed visual evidence with engineering principles to reach a conclusion. We introduce MMArch, a benchmark for architecture and civil engineering spanning ten subdomains and built entirely from figures in peer-reviewed papers. Its $1{,}212$ short-answer items are produced by a decoupled planner--writer pipeline and validated through automated screening, a blind adversarial audit, and expert review, so that answering requires perceiving the relevant evidence, identifying the governing principle, and applying it, not exploiting textual or single-figure shortcuts. Evaluating $18$ open-weight and proprietary MLLMs against a domain-expert panel, we find a wide gap: the strongest open-source model attains about $30\%$ and the best proprietary system $52\%$, while human experts reach $95\%$, more than forty points ahead. Our error analysis shows that failures concentrate in applying principles and combining evidence across figures rather than in locating it, pointing to substantial headroom for future research. Code and data are available at https://dcx-swjtu.github.io/MMArch/.

Responsible NLP practice includes a) transparency, b) ethics, and c) societal impacts. The Responsible NLP Checklist aims to push these goals, and promote responsible practice. Recently, ACL released the EMNLP 2025 Checklists to aid transparency on the current research practice, which we focus on. We curate and release the first two datasets of: a) all the checklist responses and justifications from the EMNLP 2025 Main and Finding tracks; b) checklist reference linking to paper sections. We also provide the first analysis of recent EMNLP Checklists, by examining $73,922$ responses and justifications to them. For the Main track, we find that authors isolate ethics questions of the Checklist from the paper's bulk, mimicking the trend of ethics being an afterthought. We then examine \texttt{NO} responses. We find $44.9\%$ of justifications are poor or bad-faith, being brief or empty. Then, we find significant issues with the checklist design and effort of authors, namely that $6\%$ of all checklists contained logical contradictions between parent and child responses. We also find evidence of surface compliance for responsible ethics, with $53\%$ authors dismissing potential risks or social impacts of their work, for which there should be none. We compare this to the Findings track, noticing a similar trend in both tracks. Lastly, we discuss the implications of the checklist design and provide recommendations for future checklist iterations. Including: a) enforcing a minimum word count, b) enforcing more scrutiny on the risks of appliances.

GUI agents have advanced rapidly, producing a growing body of frameworks, benchmarks, and applications. However, this growth has outpaced the maturity of the field. GUI agents remain technically brittle, incompletely engineered, and insufficiently validated for sustained real-world use. They are evolving into closed-loop software systems. Within these systems, model reasoning is coupled with interface perception, execution feedback, recovery, and human oversight. This evolution calls for a software engineering perspective that remains largely absent from existing research. We address this gap by reviewing 336 GUI-agent papers from January 2018 to April 2026. Five research questions examine the research landscape, architectures, evaluation, software lifecycle concerns, and future opportunities. Our findings show that the field has expanded sharply since 2024, while mobile and web settings remain dominant. Architectures increasingly adopt modular perceive-reason-act loops, but recovery, human escalation, safety enforcement, and auditability remain underdeveloped. This architectural imbalance extends to evaluation. Evaluations are becoming more interactive, but they remain centered on task success and are difficult to compare across protocols. More broadly, existing studies provide limited support for testing beyond benchmarks and for maintaining agents after release. Observability, privacy engineering, and systematic human oversight are also underdeveloped. Together, these findings show that capability improvements alone cannot ensure deployment readiness. Future research should connect dependable execution with lifecycle-centered testing and reproducible evaluation. It should also integrate permission and privacy controls with cost-aware, human-centered governance. This integration is necessary to build dependable, maintainable, secure, and deployable GUI-agent systems.

Verified code generation asks a large language model (LLM) to generate both an executable program and a machine-checkable proof that the program meets a formal specification, promising software that is correct by construction. The de facto workflow decouples the two halves of the problem: first synthesize a program, then attempt to prove it correct. We observe that this sequential pipeline can be both ineffective and inefficient in practice. A program generated without anticipating its proof can be subtly incorrect or structurally difficult to verify, forcing the LLM into brittle repair loops that alternate between patching the code and patching the proof. Inspired by Dijkstra's view that a program and its correctness argument should be developed hand in hand, we propose $P^3$, an LLM-based agentic workflow that first derives a unified program-and-proof plan from the specification, then elaborates the implementation and proof scaffold under this shared plan. To evaluate verified code generation in realistic settings, we further introduce Lean4Commit0, a repository-derived, library-level benchmark built by extracting core APIs from real-world software repositories and translating their requirements, including relational specifications across APIs, into Lean tasks. Using four frontier LLM backends, we evaluate $P^3$ on Verina, AlgoVeri, and our Lean4Commit0 benchmark, where it achieves the highest solve rate in every benchmark--model setting. Compared with the stronger baseline, it improves solve rates by 4.6--11.2 percentage points and reduces per-task API cost by up to roughly 40\% and wall-clock time by up to roughly 37\% on the difficult subset of each benchmark. A targeted ablation further shows gains of 3.3--8.3 points over implementation-only planning, isolating the benefit of planning the program and proof jointly.

Planetary geology relies on historical, interpretive reasoning to reconstruct past events from diverse observations. Here, we present a step toward an automated "machine intelligence geologist" by embedding this distinct methodology of geologic knowledge discovery and inference into a multimodal vision-language architecture. Focusing on the stratigraphy of lunar basaltic mare volcanism, we train a model to generate verifiably grounded geologic interpretations directly from co-registered topographic, spectral, and geologic maps. We demonstrate that while the system successfully balances established geological priors with local visual evidence to accurately describe stratigraphy and terrain, numeric age dating derived solely from vision defaults to memorized priors. Integrating an open-book retrieval mechanism resolves this, enabling the model to faithfully cite published chronologies. Our findings delineate the necessary architecture for automated geologic inference: site evidence must be visually interpreted from local data, while quantitative historical context must be retrieved from the scientific record.

LLMs have demonstrated strong capabilities in code generation and automated program repair, but migrating an entire repository rarely produces a runnable application because long-horizon translation challenges LLM-based agents' ability to maintain repository-level migration objectives. In this work, we propose Entropy-based Code Adversarial Translation (ECAT), a multi-agent framework for automated Android-to-HarmonyOS repository migration. ECAT formulates repository migration as adversarial entropy minimization through a generator-discriminator architecture. The discriminator measures migration quality using a unified metric called Code Entropy and produces text gradients that specify both file-level generation directives and the skills needed to execute them. Guided by these optimization signals, the generator iteratively updates the repository, and each update is accepted only if it reduces Code Entropy. Repeated generator--discriminator interactions progressively drive the migration from an initial template toward a functionally complete HarmonyOS repository. Successful low-entropy trajectories are further distilled into a self-evolving memory tree, enabling transferable migration knowledge across repositories. We also introduce A2H-RepoBench, the first real-world benchmark for Android-to-HarmonyOS repository migration, covering applications from tens of thousands to hundreds of thousands of lines of code. Evaluated by node alignment and an agent-based functional judge, ECAT achieves 74.7% overall migration quality and consistently outperforms existing agent-based methods across repositories of different scales.

Group-based reinforcement learning objectives such as GRPO can allocate learning signal poorly across prompt difficulty: under binary rewards, group normalization induces a divergent weighting on easy prompts. We introduce Softmax Advantage Group Estimation (SoftmaxGRPO), a drop-in alternative that replaces z-score-normalized group advantages with temperature-scaled softmax advantages, keeping weights bounded regardless of prompt difficulty. For binary rewards, we derive the exact finite-group population objective and identify MaxRL as its low-temperature limit. For bounded scalar rewards, we show that the large-group update exactly optimizes a log-moment-generating-function objective, while a universal finite-group scalar objective cannot exist without additional assumptions on the reward distribution. Empirically, SoftmaxGRPO reallocates measured gradient budget away from near-solved prompts and consistently improves over GRPO under identical rewards. It reaches 51.8% on DeepMath with verifiable rewards and improves a 1.5B instruction-tuned model from 35.0% to 68.0% on Poetry using only lightweight text-similarity rewards.

Fine-grained cross-modal understanding in drone views is essential for aerial vision-language navigation. However, the inherent wide field of view and overhead perspective of drone scenarios impose dual challenges on vision-language understanding. At the macro level, overwhelming background clutter in visual representations leads to Cross-Modal Focus Misalignment, where the model prioritizes global environmental similarities over specific object details. At the micro level, Visual Isomorphism creates ambiguity, where candidates share similar geometric structures yet differ only in subtle attributes. To address these challenges, we propose the Granularity-Aware Region Alignment and Semantic Prototype (GRASP) learning framework, enhancing discriminative capability through two synergistic strategies. Specifically, we introduce Region-Focused Alignment (RFA) to promote object-centric cross-modal alignment while suppressing background interference. Concurrently, to tackle visual isomorphism, we propose Semantic Perturbation Enhanced Matching (SPEM), which leverages a foreground-purified Semantic Prototype Codebook (SPC) to construct semantically perturbed negatives for fine-grained semantic discrimination. Extensive experiments on the GeoText-1652 benchmark and the unseen ERA dataset demonstrate that GRASP achieves competitive performance in drone-view fine-grained image-text retrieval, validating its effectiveness for cross-modal understanding in aerial scenarios. Our code implementation is available at https://github.com/UCAS-JC/GRASP.

Visual modality has recently been explored as a way to compress textual tokens, including rendering code as images for static code understanding. We study whether this representation can serve as operational context for agentic coding, where an agent must navigate repositories, edit source files, and verify executable patches. Using SWE-bench Verified, we evaluate rendered code in repository-level repair workflows and introduce controlled agent settings to separate unguided repository exploration from more structured repair stages. Our results show a mixed picture. Rendered code consistently reduces prompt-token cost, but the savings do not increase linearly with the nominal visual compression ratio. It largely preserves end-to-end repair accuracy, but does not overcome the performance limits of the underlying model or agent architecture, and can become unstable under aggressive compression. Further analysis suggests that visual code is most useful when raw source reading is a major bottleneck; once repository localization is structured, much of the remaining cost comes from patch--test trial-and-error, where visual compression has limited leverage. Overall, our study positions rendered code as a viable but conditional compression mechanism for realistic coding agents.

Medical world models commonly learn from multimodal recordings synchronized onto a fixed-rate grid. This preprocessing resamples each native stream onto a shared time axis. Each stream has an observation clock that governs when observations are emitted or updated. When this clock depends on the latent or acquisition state, it is endogenous. In such settings, synchronization may not be neutral and can erase task-relevant evidence before the model sees the data. We introduce a four-regime taxonomy that characterizes where the evidence needed to distinguish a target event or state survives. The relevant witness may remain in the sampled values, in grid-cell update patterns, in native timing, or only in an external acquisition channel. EndoClock operationalizes this taxonomy as a conservative pretraining audit. It reports the lowest witness-bearing representation supported by the available evidence, or unresolved when no regime can be established. We illustrate this failure in echocardiography, where B-mode video write-outs cease during pulsed-wave Doppler acquisition while the corresponding measurement events remain recorded only in an external acquisition log. This work is a preliminary failure alert and executable audit. Its practical message is to preserve the native observation process long enough to determine whether synchronization has erased information required by the intended task.

Outcome verifiers score completed reasoning traces but do not assign credit to intermediate tokens. Privileged self-distillation attempts to fill this gap by rescoring a model's own rollout with training-only information. A token likelihood change, however, is not automatically outcome credit. We separate three questions: whether the score tracks better actions, whether feedback construction changes what is compared, and what behavior the training loss reinforces. We establish these distinctions formally. When a rollout is scored using hindsight feedback written about that same rollout, its content determines both the tokens and the scoring context, creating direct self-dependence. Using feedback from another rollout of the same problem removes this dependence but does not guarantee a useful score. In matched experiments with a 20B model on AIME 2025, the implemented additive score is near chance (AUC=0.505) and slightly favors incorrect traces after length adjustment. In the paired comparison, the outcome-only control records 64.2\%, versus 24.2\%--33.9\% for five token-score variants. The results motivate validating score meaning, feedback construction, and training behavior separately before calling a likelihood signal credit.

Large language models (LLMs) have advanced Text-to-SQL by enabling natural language interfaces to databases without task-specific fine-tuning. However, existing LLM-based systems remain unreliable, often generating SQL queries that are invalid under the database schema, referencing non-existent tables, attributes, functions, or values. Such errors persist because interactions with the database management system (DBMS) are typically limited to error messages, leaving it in a largely passive role during query refinement. This paper proposes SafeQL, \textit{a search-based refinement paradigm that redefines the role of the DBMS as an active guide in the refinement process}. Instead of regenerating entire queries after execution failure, SafeQL interprets DBMS feedback to incrementally repair only the erroneous components. Each refinement step is formulated as a guided search within a \textit{safe query space}, where candidate queries are progressively validated through DBMS execution, thereby converging to an executable query and preventing repeated regeneration of errors. Experiments on the Bird and Spider benchmarks show that SafeQL significantly improves execution accuracy and efficiency compared to regeneration-based methods.

Residential energy estimates are often needed before detailed envelope characteristics, equipment efficiencies, infiltration, sensor, or billing data are available. This study quantifies the trade-off between predictive accuracy and input accessibility using two nationally representative U.S. residential-energy datasets: the survey-based Residential Energy Consumption Survey (RECS) and the simulation-based ResStock dataset. Full-feature models were first used to establish dataset-specific performance benchmarks. For total-energy estimation, the models were subsequently restricted to ten low-burden variables obtainable from occupants, administrative records, or location-based weather data without an on-site energy audit. Among CatBoost, XGBoost, LightGBM, Random Forest, and Neural Networks, CatBoost consistently achieved the highest predictive performance for the full-feature analysis, reaching R2 = 0.90 for ResStock and R2 = 0.73 for RECS. When the feature set was restricted to ten homeowner-accessible inputs to simulate realistic deployment conditions, model performance converged to R2 = 0.61 for RECS and R2 = 0.62 for ResStock, showing that algorithmic complexity cannot fully compensate for missing physical and behavioral information. However, for a more homogeneous ResStock cohort consisting of single-family detached, natural-gas-heated homes in Climate Zone 6A constructed between 2000 and 2010, a reduced-input model improved accuracy to R2 = 0.85, demonstrating the value of targeted modeling for homogeneous populations. The results indicate that tree-based ensemble models can serve as high-fidelity emulators of national-scale residential energy datasets. However, careful consideration of feature availability, dataset origin (empirical vs. synthetic), and applicable use cases are also important.

LLM analytics agents are evaluated on SQL syntax accuracy, but production failures look different: questions with two valid business definitions, questions the warehouse cannot answer, deprecated columns after a schema change, and queries that execute successfully while returning the wrong business number. No execution-match metric can score them. This paper introduces WarehouseReliabilityBench, 400 frozen tasks over two synthetic warehouses in which roughly half the correct responses are a clarification, an abstention or a refusal, with pinned denominators and a pre-registered paired bootstrap fixing each claim verb before the numbers existed. QueryProof, a 7B agent, uses rules derived from a semantic layer and physical catalog to determine its behaviour, and gates every answer on deterministic post-execution checks. On an 80-task synthetic test split evaluated once, QueryProof outperforms a direct-prompted 32B baseline by +0.237 [+0.112, +0.375] Business Truth Rate at 71.0% lower cost per correct answer; against a cost-matched few-shot baseline the accuracy gain holds but the cost difference does not resolve. This compares systems rather than model sizes: the 32B baseline receives none of the scaffolding. False success falls from 0.754 to 0.351 of returned answers, and no wrong number was returned on an answerable task (0 of 24), though 13 answers went to questions requiring clarification or abstention. Removing the routing layer changes little (0.562 against 0.537), so the result does not depend on escalation. Routing tuned on validation over-abstains on test, and the fitted confidence model loses to the heuristic it replaced. Resampling template families rather than tasks widens both accuracy intervals to include zero, so the effect's direction is better supported than its magnitude. The gain tracks the deterministic layer, though no component ablation was run.

LLM agents are increasingly equipped with skills to perform complex tasks through multi-step reasoning and tool use. Although skills provide reusable procedural knowledge, agents may still execute them unreliably. Even when an agent has demonstrated the capability to complete tasks under the guidance of a skill, it may fail to do so consistently across similar tasks or repeated runs due to deviations from the skill procedure or incorrect execution of individual steps. Such instability limits the practical reliability of LLM agents. To address this problem, we propose SkillSentry, a skill-oriented runtime assurance framework built upon a new domain-specific language (DSL) for representing runtime guidance for skill execution. SkillSentry initializes the runtime guidance by combining a skill specification extracted from the corresponding skill document with execution experience mined from historical successful and failed traces. It then wraps around the agent execution loop to monitor and guide skill execution under the current guidance, while iteratively refining the guidance using newly collected traces. We evaluate SkillSentry on 15 skills across two LLM agents, each paired with two backbone models, i.e., Claude Code with Claude-Haiku-4.5 and Claude-Opus-4.6, and Codex with GPT-5.2 and GPT-5.4. Our results show that SkillSentry improves the task success rate of LLM agents by 24.1% across skills, on average, while exhibiting lower variability across repeated runs.

Large language model-based multi-agent systems have recently shown strong potential for complex, long-horizon tasks. However, existing methods mainly rely on coarse prompt-level differentiation without parameter adaptation for diverse subtasks, resulting in insufficient inter-agent heterogeneity and limited specialized capability that bottleneck performance on tasks with complex requirements. To address this, we introduce a Task-Oriented Multi-Agent System with Mixture of Role-Subtask Experts (MoRSE) that distinguishes agents with (role, subtask)-conditional specialization at both the task structure and parameter levels. To make agents' responsibility explicit at the task structure level, we formulate a task-oriented multi-agent system that decomposes each task into a dependency-aware Directed Acyclic Graph of subtasks and assigns each agent a specific (role, subtask), introducing task-level specialization across collaborating agents. Additionally, to address the diverse role and subtask parameter adaptation demands, we propose a dynamic Mixture of (role, subtask) LoRA Experts module with a prototype-based semantic router for subtasks, augmenting agents with parameter-level specialization on a shared LLM substrate cost-effectively. Then, to co-optimize experts and router stably under sparse task rewards, we further propose a hierarchical group-relative policy optimization with two-layer credit assignment that isolates expert updates from the cross-route variance introduced by routing decisions, disentangling expert quality from routing quality. Experiments on code-generation benchmarks across three backbones demonstrate the effectiveness of our approach, with improvements in both whole-task and step-wise performance, and the gains from trained specialization generalize across held-out task categories and domains.

Federated learning (FL) must serve devices with varying computational capabilities. A fixed model cannot suit all devices, while training one model per deployment limit is costly. Federated supernet training instead learns one elastic model with differently sized subnetworks, then deploys a suitable one to each device. When client inference budgets differ, however, parameters exclusive to high-cost subnetworks are reachable by fewer clients. We propose FEAST, a federated shared-space training framework that counters this imbalance by jointly training multiple subnetworks within each client's limit. Budget-tailored sub-supernet routing sends only the relevant supernet portion, and sparse aggregation merges the returned parameter slices. The trained supernet directly serves the subnetworks used during federation and supports post-hoc extraction of additional subnetworks without federated retraining. We further show that independently assigning clients' training-data volumes and inference budgets can distort accuracy--inference-cost comparisons in heterogeneous FL simulations, and introduce a one-parameter $γ$-allocation protocol to control this coupling. In our experimental setup, the SuperFedNAS and DeepFedNAS supernet training procedures remain near chance at 25M and reach at most $17.09\%$ at $596$M inference MACs; FEAST reaches $71.06\%$ at $596$M, $2.4$ points above the strongest model-heterogeneous weight-sharing baseline at its largest tier. Across CIFAR-100, CINIC-10, and TinyImageNet-200, FEAST achieves the highest population-averaged accuracy among the evaluated weight-sharing methods when each client receives its largest affordable subnetwork. Sub-supernet routing reduces aggregate model-parameter traffic by $6.8\times$ relative to full-supernet transmission.

Skill-based LLM agents select reusable procedures from an external library to solve complex tasks, yet their routing decisions rely entirely on text-level signals such as task descriptions, verbal reflections, and experience-derived rules, while the model's own internal representational state remains unobserved. Recent interpretability work has shown that LLMs maintain linear emotion representations that causally influence behavior; however, these representations have been exploited only for post-hoc analysis or direct output steering, and have not been used to inform agent-level decision-making. We propose Emotion2Skill, a framework that extracts LLM-internal emotion vectors and incorporates them into both skill selection and skill evolution. At each decision step, a 27-dimensional emotion state is extracted from the residual stream and mapped to a confidence-gated summary injected into the routing prompt. Beyond online selection, emotion trajectories are analyzed for abrupt internal-state shifts to pinpoint problematic skill invocations, guiding targeted SOP rewriting that replaces the coarse binary outcome signal of prior methods. On WebShop and ALFWorld, Emotion2Skill with Qwen3-8B improves over the Zero-Shot baseline by +26.9% success rate and +25.5% average success respectively, outperforming all baselines on both benchmarks with consistent gains on Qwen3-14B. Co-activation analysis further reveals semantically coherent emotion--skill pairings, confirming that the routing improvements reflect meaningful internal-state signals rather than opaque statistical correlations. These results establish LLM-internal emotion representations as an effective decision-level signal for orchestrating agent skill systems, extending their utility beyond interpretability and output steering. The code is available at https://github.com/BoHan-LIN04/Emotion2Skill.

Industrial processes are complex systems composed of multiple interacting sensors that generate multivariate time series (MTS). Detecting anomalies in such systems is critical for reliability and safety, yet understanding their origin is equally important. Existing Graph Neural Network (GNN)based methods for anomaly detection primarily focus on sensor-level deviations and either attribute anomalies directly to the deviating sensors. When diagnosis is attempted, generally, the most deviated sensor is identified as a root cause of a system fault. However, in many industrial systems, anomalies do not arise from faulty sensors but from disruptions in the influences governing the system dynamics. We propose an explainable GNN-based anomaly detection framework that shifts the perspective from sensor-level anomalies to component-level diagnosis, hypothesizing that anomalous measurements are symptoms of altered inter-sensor influences. Experiments show that the method effectively identifies and prioritizes the true faulty components, providing interpretable insights into system failures.

Multimodal federated learning (FL) supports collaborative modeling in privacy-sensitive health-sensing and medical settings, but realistic deployments often exhibit dual-axis modality missingness: clients have different modality sets, and individual samples may contain only subsets of the modalities available locally. Existing methods typically address these two axes separately. We propose Flux, a multimodal federated learning framework built around two complementary components. First, modality-aware confidence tempering learns sample-specific confidence for each modality through mask-aware unimodal supervision and fuses the confidence estimates from observed modalities into a sample-adaptive temperature that adjusts predictive sharpness according to evidence quality and completeness. Second, gradient-decoupled private adaptation applies this temperature only to a client-private prediction pathway, while training the shared federated model with a standard, untempered objective. This enables sample-specific, client-local confidence adaptation without allowing confidence-dependent gradients to perturb shared representation learning. Across four multimodal datasets, Flux achieves the highest average macro-F1 on every dataset, outperforming the strongest dataset-specific baseline by 0.8~2.2 points and by 1.6 points on average. Additional analyses demonstrate favorable calibration, temperature sensitivity to both modality missingness and input corruption, and more stable shared optimization under private-only tempering. Our code is available at https://github.com/AdibaOrz/Flux.

Federated learning enables privacy-preserving collaboration across distributed devices without centralizing local data. However, clients may differ not only in data distributions but also in domain knowledge and annotation capabilities. In this paper, we introduce label granularity skew, a new form of statistical heterogeneity in federated hierarchical classification, in which clients provide taxonomy-consistent labels at different levels of detail within a shared class hierarchy. To model this heterogeneity, we generate client-specific local label hierarchies using a probabilistic relational neighbor classifier and construct a WordNet-guided hierarchy via silhouette score-based coarsening. Our analysis shows that strongly coupled hierarchical models are sensitive to incomplete supervision, while the conditional softmax classifier is more robust. Based on this insight, we propose Branch-wise Decoupled Fine-Tuning (BDFT) and its federated version, FedBDFT, which fine-tune branch-wise classifiers and aggregate them through federated optimization. Experiments on CIFAR-100, TinyImageNet, and ImageNet show that FedBDFT substantially improves robustness under severe label granularity skew, with average gains of 27.9% and 56.4% at skewness levels of 0.6 and 0.9, respectively. Zero-shot results further indicate that FedBDFT better preserves hierarchical representations for unseen fine-grained classes. These findings demonstrate its effectiveness for federated hierarchical classification with heterogeneous label granularities.

Flow-matching models are now a mainstream method to image generation, but its adaptation to diverse downstream scenarios typically relies on post-training, which may cause conflicts among task-specific optimization objectives. Reinforcement learning enables direct optimization of task-specific rewards beyond the original models, yet trajectory-level optimization may incur high-variance gradients and cross-task interference. On-policy distillation (OPD) offers dense and stable supervision on student rollouts, but conventional teacher matching remains imitation-based. We propose DreOPD, a Degraded-reference extrapolative OPD method for flow-matching models that bridges these two paradigms. Our DreOPD converts implicit reward extrapolation into closed-form velocity regression, enabling extrapolative post-training with the stability of OPD. It further uses a mildly degraded reference to strengthen the teacher-reference contrast, yielding a clearer extrapolation direction. Experiments on single- and multi-teacher settings show that DreOPD outperforms OPD and multi-task RL baselines in average performance, while surpassing specialized teachers on most metrics.

Industrial-safety understanding requires more than detecting workers, equipment, and personal protective equipment. Models must also assess compliance, identify hazardous interactions, explain potential accident mechanisms, and recommend preventive actions. Existing safety datasets primarily focus on visual perception or isolated violation recognition and provide limited supervision for evidence-grounded reasoning. We introduce SafeSceneReason, a multimodal industrial-safety reasoning benchmark and companion training corpus that connects workplace scenes with knowledge from occupational accident investigations. SafeSceneReason combines two complementary data-construction pipelines. The scene-centric pipeline converts annotated workplace images into executable safety scene graphs and generates deterministic answers through program execution over objects, relations, and safety rules. The report-centric pipeline extracts figures and contextual evidence from accident reports and constructs multimodal questions using evidence graphs, explicit information boundaries, multi-step reasoning paths, and iterative verification. The resulting resource contains 110,581 verified scene-centric question--answer pairs and 13,114 refined report-centric question--answer pairs, covering perception, spatial and quantitative reasoning, compliance assessment, evidence synthesis, causal analysis, and mitigation-oriented decision making. Evaluation of representative proprietary and open-source vision--language models reveals substantial performance differences and persistent weaknesses in comparative, technical, and multi-evidence reasoning, demonstrating that strong general visual understanding does not yet guarantee reliable industrial-safety reasoning.

On-Policy Self-Distillation (OPSD) is commonly interpreted as the transfer of privileged information: a teacher observes the verified solution to the target problem and supervises the student's trajectory. However, this interpretation conflates two effects. The reference solution not only reveals the answer to the current instance but also changes the context under which the teacher provides token-level supervision. We investigate the role of target-specific privilege with $\mathrm{OP}^{2}\mathrm{SD}$ (On-Policy Self-Distillation from Other Problems), which replaces the paired reference with a problem and solution from a different example, while preserving the student rollout, teacher, and distillation objective. Across three models and three mathematics benchmarks, $\mathrm{OP}^{2}\mathrm{SD}$ improves over the base model, remains competitive with OPSD. The success of $\mathrm{OP}^{2}\mathrm{SD}$ implies that OPSD gains do not necessarily come from access to the reference solution, and that the teacher's context-induced behavior is an important factor.

Omnimodal language models (OLMs) enable unified audio-visual understanding, but processing long joint token sequences makes inference computationally prohibitive. While recent token compression methods attempt to alleviate this burden, compressing modalities in isolation often destroys the temporal cross-modal anchors necessary for coherent reasoning. We introduce Omni2LoRA, a two-stage framework for efficient parametric memory compression via coherence-preserving context distillation that bypasses the token bottleneck entirely. First, a Perceiver hypernetwork processes intermediate representations from a frozen OLM to encode the multimodal context into a full-rank Low-Rank Adaptation (LoRA) adapter in a single forward pass. To prevent the resulting parameter footprint from scaling linearly with recording length, we optimize a discrete rank allocation policy via Group Relative Policy Optimization (GRPO) that uses a modality-ablated counterfactual reward to explicitly penalize the loss of audio-visual coherence, forcing the model to allocate its fixed sub-linear rank budget to synergistic cross-modal anchors rather than isolated visual features. Across three omnimodal backbones, Omni2LoRA operating at a 30% rank budget outperforms direct full-context inference and strong token-compression baselines (OmniZip, OMAC, O-MARC) on four audio-visual question answering benchmarks, improving average accuracy by 8-12% over the strongest baseline and remaining stable under compression ratios as tight as 75%, where token-pruning methods degrade sharply. By converting multimodal memory into a fixed-budget, reusable parameter state, our method drives answer-time multimodal-token load to zero, cutting per-query Time to First Token (TTFT) by up to 12x relative to full-context inference and amortizing to under 0.5s after a handful of queries.

Efficient text-to-image generation requires both reinforcement-learning (RL)-based reward alignment and few-step distillation, yet these procedures are typically performed sequentially, increasing training cost and risking the loss of reward gains during compression. We instead take an RL-native perspective: diffusion RL already generates reward-scored finite-step trajectories, whose intermediate states provide a natural source of distillation supervision rather than a disposable byproduct of sampling. Based on this insight, we propose REST (Reward-Enhanced Scored-Trajectory Distillation), a single-stage RL-distillation co-training framework that attaches a decoupled student to an arbitrary RL teacher. The student learns segment-wise from the teacher's evolving rollout trajectories while leaving the original teacher optimization unchanged. To prevent uniform imitation from preserving undesirable low-reward behaviors, we further introduce Advantage-Modulated Distillation (AMD), which transforms rollout advantages into signed weights over a base distillation loss. AMD strengthens supervision from preferred trajectories and mildly repels the student from low-reward ones. The resulting framework is lightweight and plug-and-play, requires no extra image rollouts, no separate distillation dataset, and no adversarial training. Experiments on compositional generation, visual text rendering, and human-preference alignment show that REST enables few-step CFG-free inference that matches or surpasses its 40-step RL teacher, with an overall additional training cost below 25% over pure RL. REST improves DrawBench PickScore over RTDMD by 0.82 while requiring only one-fifth of the training iterations.

The key-value (KV) cache is the primary throughput optimization in modern large language model (LLM) inference, enabling prefix reuse across requests. In multi-tenant deployments this cache is shared across tenants, creating a timing side channel: an adversarial tenant can reconstruct another tenant's private prompt by probing cache-hit latency. Three published attacks exploit it -- PROMPTPEEK, EarlyBird and InputSnatch -- reaching up to 100% attack success rate against unprotected vLLM and SGLang, with rates varying by cache architecture and prompt structure. We present KVGov, a governance layer addressing all three attack families' prefix-cache paths under one mechanism. A per-principal salt sigma_p = HMAC_K(secret, principal_id) seeds the block-hash chain, making cache keys cryptographically disjoint across principals. An ablation (N=1000 trials, seed 2026, deterministic judges) isolates this salt as the necessary and sufficient component. KVGov adds ORIGAMI, a Stackelberg water-filling audit scheduler that reduces adversary expected utility by 12.6% at realistic tenant heterogeneity (Gini 0.63), and an evolutionary stability analysis giving a 31.6% adversary-prevalence tipping point below which global caching remains stable. On real hardware (Qwen2.5-7B-Instruct, vLLM 0.26.0, NVIDIA A100) we measure a gate-verified cold/cached TTFT ratio of 0.22, confirming the channel is exploitable at production scale; the defense itself is evaluated in simulation calibrated to those measurements. We replicate the channel on an independent stack (llama.cpp on Apple Metal, ratio 0.093). Finally, isolation and cache efficiency need not conflict: identifying information resides only where prompts diverge, so injecting the salt at that boundary rather than the chain root retains an estimated 93% of the prefix-cache benefit with no cross-principal signal.

Inverse decision modeling infers latent properties of decision processes from observed behavior, but existing formulations rely primarily on action trajectories. In verbalized cognitive tasks, task execution also produces response dynamics that action-only formulations leave unmodeled, such as verbal production, interaction, and hesitation. We propose a factorized inverse decision model (FIDM) that decomposes each individual's task-execution likelihood into an action factor and an effort factor, governed by separate individual-specific parameters. From raw verbal transcripts, a language model produces structured task-execution traces for factorized inference. On data from 400 older adults performing a grocery-shopping dialog task for cognitive screening, controlled recovery shows selective estimation of the intended factors, while matched semi-synthetic conditions show that FIDM preserves action-execution distinctions even when aggregate behavioral summaries are matched. Action evidence further localizes task-defined deviations across participants. In cognitive-status classification, FIDM provides information complementary to clinical scores, trajectory summaries, and frozen language representations, with consistent gains across all evaluated baselines in the binary setting.

Federated Learning (FL) enables collaborative model training across distributed client devices while preserving data privacy. However, FL faces significant challenges due to data heterogeneity, particularly in terms of label distribution skewness and variations in dataset sizes, which can lead to biased model updates and hinder convergence. To address this, we propose FedTVD, a novel FL algorithm that weights client contributions during aggregation by considering both data quality and quantity. Unlike traditional FL approaches such as FedAvg, which rely solely on dataset size for client weighting, FedTVD integrates Total Variation Distance (TVD) to measure the divergence between each client's local label distribution and a uniform global distribution. Clients with highly skewed distributions receive lower weights, preventing unbalanced datasets with imbalances from disproportionately influencing the global model. At the same time, dataset size is incorporated to ensure scalability and fairness. This dual-weighting mechanism effectively mitigates the impact of data imbalance, leading to more stable and generalized global models. Experimental results show that FedTVD consistently outperforms state-of-the-art methods across all datasets (FMNIST, CIFAR-10, and CIFAR-100) and all levels of data heterogeneity. Notably, it achieves up to 10.6% improvement over FedAvg on CIFAR-10 under highly skewed data, while maintaining top performance even under moderate and IID settings.

Score-driven filters multiply a scaled log-likelihood score by a gain that controls the update magnitude. We treat this gain as a decision variable and study its online learning. Conditional on the current state, observation, score, and scaling rule, each admissible gain induces a reachable next state and a one-step-ahead predictive density: scalar gains govern distance along a line, while diagonal gains govern coordinatewise transmission. Gain selection is therefore a conditional predictive decision problem with a Kullback-Leibler objective. For a scalar unscaled gain, the negative raw product of consecutive scores is the stochastic gradient of this loss; positive aGAS scaling only rescales the effective step. Monotone differentiable gain links induce mirror-descent geometries on bounded gain domains, while persistence yields a Bregman pull towards a reference gain. Under convexity, compactness, and regularity conditions, we establish dynamic-regret bounds for projected and discounted mirror updates relative to time-varying, current-information comparators. Simulations illustrate the roles of scaling, link geometry, persistence, and coordinatewise transmission rates. An out-of-sample panel of equity-index volatilities shows that the bounded mirror gain generally matches or outperforms a constant gain while avoiding the extreme spikes of a nominally unbounded exponential link, with the strongest improvements observed in multi-crisis markets.

Reinforcement learning (RL) has become a central post-training paradigm for eliciting reasoning capabilities in large language models, yet uniform task sampling allocates compute without regard to differences in how tasks respond to optimization. Existing task-valuation methods mostly rely on snapshot-based signals such as current pass rate or reward, which estimate how solvable a task is under the current policy. However, tasks with similar current solvability can still differ substantially in how positively they respond to further training. We study this residual axis as task learnability: a regime-conditional measure of expected positive response to continued training under a fixed RL post-training regime. By analyzing per-task reward trajectories, we find that learnability is reproducible across independently sampled training contexts and predictive of downstream utility. To make this signal practical before training begins, we propose TrajVal, a lightweight probe-based estimator that approximates per-task learnability from a short probe run and two endpoint evaluations. TrajVal can be used either as a standalone static prior for task sampling or as a multiplicative prior for existing online schedulers. Experiments on mathematical and logical reasoning benchmarks across multiple model scales show that TrajVal improves data efficiency over uniform sampling and provides complementary gains when combined with online scheduling methods.

Recent Retrieval-Augmented Generation (RAG) systems increasingly combine vector retrieval with structured knowledge, such as Graph RAG and Filtered vector search. However, existing database architectures struggle to support such complex RAG workflows efficiently, as they rely on out-of-DB pipelines or in-DB non-native integration, leading to high overhead. This demo paper presents AkasicDB, a database system that natively supports such RAG workflows by jointly executing vector similarity search, graph traversal, and relational filtering within a single execution framework. AkasicDB extends our prior work, Chimera, with native vector support to enable such unified execution. Based on AkasicDB, we demonstrate the first native integration of Vector-Graph-Relational RAG, which we refer to as Omni RAG. Through an interactive chat-style demonstration, users execute and visualize Omni RAG queries, directly experiencing its superior retrieval and reasoning over vector-only approaches while observing the practical limitations of existing database architectures in supporting Omni RAG. A demonstration video is available at https://youtu.be/8d09_dtrEIM

Neural language models trained on large crowdsourced corpora frequently exploit spurious surface patterns tied to target labels without true linguistic or causal relevance, boosting benchmark performance while failing on adversarial or out-of-distribution inputs. Existing approaches either require manual specification of the feature vocabulary or automate discovery only partially, leaving the gap between dataset-level correlation and model-level exploitation unaddressed. We present U N M ASK, a fully automated pipeline that discovers, causally verifies, and mitigates spurious correlations in text classifiers without additional human annotation. Given unlabeled training examples, U N M ASK generates candidate surface patterns as executable boolean expressions, filters them through a statistical validation protocol with independent replication, and establishes causal model dependence via verified counterfactual interventions. Causally confirmed features then serve as annotation-free group definitions for Deep Feature Reweighting, eliminating the group labels that standard DFR requires. Applied to BERT and RoBERTa trained on MNLI, our pipeline independently rediscovers established lexical-overlap and negation biases, verifying 9 of 10 features on BERT and 6 on RoBERTa, and improving HANS accuracy by up to 12.58 pp. On CivilComments-WILDS, programmatic groups match the 70.1% worst- group accuracy of hand-labeled DFR (Kirichenko et al., 2023) without demographic annotation. We further demonstrate that the discovery and validation stages generalize to reward model preference data, surfacing interpretable spurious correlations in RewardBench2.

Decentralized intelligence systems with heterogeneous devices and limited coordination increasingly rely on decentralized federated learning (DFL). However, DFL suffers from convergence inefficiency under data heterogeneity due to the use of a uniform learning rate (LR) that ignores layer-specific optimization needs. Foundational layers are responsible for maintaining network consensus, while specialized layers adapt to local data characteristics, leading to conflicting gradients and degraded performance under non-IID conditions. To address this fundamental tension, this work introduces FedA2L, a method that dynamically adjusts layer-wise LRs based on model divergence signals. By leveraging local update intensity and network consensus constraints, FedA2L seamlessly integrates into existing DFL protocols without additional communication or coordination. Extensive evaluations across DFL algorithms, various model architectures, and datasets demonstrate that FedA2L achieves up to 4.94 times faster convergence than vanilla DFL and reduces communication rounds by up to 59% compared to scheduler-based baselines. Furthermore, FedA2L exhibits resilience to severe data heterogeneity, larger network sizes, and sparse topologies, reducing communication overhead and establishing it as a versatile optimization tool for resource-constrained or large-scale distributed learning in edge and IoT deployments. The code is released at https://github.com/nclabteam/FedA2L.

Modern autonomous vehicles are equipped with multiple sensors, such as cameras, LiDAR, and radar, for comprehensive environmental perception. However, robust cross-modal feature fusion remains a critical challenge, as the reliability of each sensor varies significantly across diverse real-world driving conditions, including poor visibility and adverse weather. While uncertainty quantification (UQ) mitigates this issue by allowing models to prioritize reliable signals, existing uncertainty-aware fusion methods typically rely on simple feature-level uncertainty estimates and thus often fail to generalize effectively in complex, out-of-distribution scenarios. To address this limitation, we propose CRUISE, a novel uncertainty-aware cross-modal sensor fusion framework. CRUISE integrates a vision-language model (VLM)-guided UQ module that generates fine-grained, pixel-level uncertainty estimates. By leveraging the VLM's rich prior knowledge and superior contextual reasoning, our approach provides a highly informative guide for the fusion process. Furthermore, we introduce a dynamic adaptive mechanism that explicitly models and captures cross-modal dependencies, ensuring the framework fully exploits the inherent complementary nature of multi-sensor inputs.

Dense expert merging combines domain-specialized language models into one single checkpoint, typically by admitting task-vector support in weight space. However, this admission is governed by three decisions that existing methods answer only partially: where to open layer capacity from cross-expert conflict, who should occupy that capacity based on domain demand, and how to admit the resulting support without relying on costly recipe search. To tackle these issues, we propose SigMerge (Signature-Guided Capacity Occupancy), a structured capacity assignment framework for dense expert merging. Starting from a dense base merge, conflict signatures set each layer's capacity from cross-expert conflict, positive base-merge deficits set each domain's share of that capacity, and a sequential occupancy rule admits each expert delta up to the resulting layer-domain budget. Across 21 paired settings spanning seven dense base merges and three model pools, SigMerge improves every one (by 15.0% on average) and achieves the best average rank (1.67) among six merging methods, outperforming three categories of merging baselines.

Unsupervised time-series domain adaptation (DA) addresses the challenge of transferring a classifier from a labeled source domain to an unlabeled target domain under distribution shifts induced by different users, sensors, devices, acquisition conditions, or temporal dynamics. Existing methods typically mitigate this shift by aligning marginal feature distributions through adversarial training, optimal transport, or moment-based discrepancies. In this paper, we propose Class-Conditional Path Distribution Alignment (CPDA), a non-adversarial discrepancy-based framework that aligns source and target class-conditional latent path distributions rather than only global feature marginals. CPDA introduces a composite signature-spectral kernel that jointly captures pooled semantic features, temporal path structure, frequency-domain information, and low-rank path-signature dynamics, while using source labels and target soft pseudo-labels to perform class-preserving alignment. We further provide a theoretical analysis showing that CPDA defines a valid kernel discrepancy, admits existing moment-matching methods as restricted cases, and yields a class-conditional target-risk bound. Extensive experiments with CNN, ResNet18, and TCN backbones on 13 different time-series DA benchmarks demonstrate the effectiveness of CPDA against 30 discrepancy, adversarial, and pseudo-labeling baselines.

GK is a query-directed first-order prover that extends ordinary resolution-based proof search with explicit positive and negative claims, numerical confidence values, and prioritized default rules with exceptions. It works directly with non-ground clauses, including equality and function terms. Candidate proofs are found by bounded first-order proof search; exception conditions of defaults are checked by further bounded searches, recursively when exceptions themselves depend on defaults. This avoids requiring a finite global grounding, while allowing incomplete searches to be reported as such. This paper adds structure-preserving quantitative reporting to that framework. Retained proof histories are used in two calculations. The first reconstructs the uncertain ground premises used by each proof and computes the probability that at least one retained proof is available, without counting shared premises independently. The second resolves positive and negative support at intermediate atoms before that support is propagated through later rules; the same calculation evaluates uncertain exception conditions for individual rule applications. Reports separate positive support, negative support, conflict, and ignorance and identify detected incomplete calculations or fallbacks. The implementation performs bounded reconstruction and dependency traversal after proof search and still requires no global grounding. Analytic examples and independent simulators reproduce the reference calculations on their stated fragments. Comparisons with probabilistic logic, probabilistic ASP, default logic, and goal-directed ASP identify cases of agreement, semantic difference, unsupported translation, and incomplete computation.

Despite significant advances in instruction-following and auditory comprehension, the evaluation of Emotional Intelligence (EI) in Spoken Language Models (SLMs) remains confined to rudimentary paralinguistic perception, lacking a systematic, theory-driven cognitive framework. We introduce EmoSBench, the first comprehensive EI evaluation benchmark for SLMs constructed upon the four-branch theoretical model, covering Perceiving, Understanding, Using, and Managing Emotion across ten sub-tasks. Preliminary assessments on EmoSBench reveal a substantial gap: even leading proprietary models like GPT-4o-Audio achieve only 52.6%, significantly trailing human baselines. To bridge this gap, we develop EmoS, a specialized evaluator model optimized via Supervised Fine-Tuning (SFT) and Group Relative Policy Optimization (GRPO). To facilitate its effective training, we curate EmoDialogue, a bilingual dataset providing necessary fine-grained supervision through response pairs with rigorously defined EI gradations. Concurrently, we introduce a reward mechanism integrating a Steep Exponential Accuracy Reward (SEAR) and a Rationale Fidelity Reward (RFR) to enforce precise ordinal scoring and valid reasoning. Experiments demonstrate that EmoS reaches 83.8% accuracy, approaching human-level performance. Furthermore, evaluations on authentic, unconstrained spoken interactions validate its robust real-world generalization, establishing a foundational framework for advancing emotionally intelligent dialogue systems.

A long-form translation request can succeed at the API layer and still produce an unusable result. The output may be empty, truncated, filtered, dominated by source or prompt material, or interrupted after producing text worth keeping. This report describes a recovery protocol developed for a deployed translation system with heterogeneous inputs and provider APIs. It delays the first visible release behind a 64-character window, validates the assembled output, and uses typed stream events to distinguish replacement from continuation. Interrupted work is retained only when a paragraph or sentence prefix can be re-derived from the source. Further attempts follow a stable model order and a shared deadline before entering a provenance-marked fallback path. A sanitized companion artifact implements the protocol and passes 38 public tests. Its fixed cases reproduce all 14 configured completion labels, contain four early-invalid prefixes before any of their 235 characters become visible, retain 31 boundary-safe characters across four interrupted streams, and satisfy the attempt, event, and provenance rules in two end-to-end scenarios. These results are executable checks of the published control flow. Translation quality and detector performance on naturally occurring outputs require a different evaluation.

Enterprise data warehouses (DWs) support business-critical analytics, but warehouse task delivery remains a complicated production process involving context retrieval, workflow configuration, code generation, platform submission, and failure diagnosis. Although large language models (LLMs) and coding agents have improved software development, they are insufficient for production DW delivery, which requires dependency-aware orchestration, lifecycle-aware artifact control, and continuous adaptation to evolving platform practices. We present SiriusDeliver, an end-to-end delivery automation agent for production warehouse task submission. SiriusDeliver integrates three components: a hierarchical delivery agent that orchestrates warehouse skills, an artifact lifecycle control module that verifies and revises artifacts before and after platform execution, and a trace-driven skill evolution mechanism that maintains reusable skills from delivery trajectories. We evaluate SiriusDeliver through offline datasets and large-scale production deployment on Tencent Cloud WeData. Offline experiments on real-world warehouse delivery cases show that SiriusDeliver improves delivery success and automation efficiency over representative baselines. During a two-month deployment across 6 business teams and 4 warehouse task types, SiriusDeliver served 3,600 monthly active users and supported 18,240 delivery sessions, achieving an 87.2% end-to-end success rate and a 73.5% autonomous submission rate. A one-month A/B test shows that SiriusDeliver reduces median delivery time from 228 to 23 minutes and engineer effort from 95 to 11 minutes, while maintaining comparable final delivery success.

Large language model (LLM) agents provide a promising interface for command-line-based network operations, but a plausible command may still fail or introduce operational risk after execution. Existing approaches mainly focus on command generation or final configuration correctness, and do not use execution-grounded experience to jointly improve candidate coverage and action selection. We propose an execution-grounded dual-path consequence-aware agent for CLI-based SONiC operations, which generates multiple complete actions, predicts their execution consequences, and selects the final action through utility- and risk-aware reranking. The proposal-side path abstracts reusable operational lessons into retrievable guidance to improve feasible-action coverage without modifying the proposal LLM, while the selection-side path adapts the consequence predictor through session-level LoRA updates using real SSH feedback to improve conditional selection quality. Experiments over multi-turn SONiC operation sessions with different Qwen3 proposal models show that the framework improves feasible-action coverage and top-1 execution success, and that the two adaptation paths provide complementary gains over interaction.

Accurate landmark localization in medical images is a fundamental step for quantitative clinical measurement and downstream analysis. Existing localization methods have advanced, among which multi-stage refinement is a superior solution. Although this strategy mitigates the anatomical ambiguity inherent in single-stage global predictions, its high computational cost limits practical applicability. In this work, we propose a parameter-economic model, PPOC-LL, which leverages Prototype learning-based Progressive Offset Correction for Landmark Localization. Our contribution is three-fold. First, to drive coarse-to-fine landmark optimization, we introduce a multi-scale dynamic perception strategy for patch-level feature pyramid modeling. Second, to effectively handle anatomically similar patterns, we design a similarity-driven prototype learning mechanism that captures informative local semantics for robust offset prediction. Last, to stabilize the model learning and improve the overall performance, we incorporate a novel error-aware reliability regularization via tolerance-based balancing. We collected a large validation cohort, including two public and one private datasets spanning X-ray and ultrasound modalities, covering cephalometric, symphysis-fetal head, and fetal heart landmarks. Extensive experiments demonstrate that PPOC-LL achieves satisfactory performance with a favorable trade-off between accuracy and model complexity.

Static Application Security Testing (SAST) tools have become indispensable in modern secure software devel- opment. However, these tools often generate false-positive (FP) alerts, imposing substantial manual inspection costs and reducing the trust from developers. Existing FP reduction methods still face two primary challenges. First, the large differences among SAST tools and vulnerability categories make it difficult for these methods to learn recurring patterns in historical false positives. Moreover, the knowledge used by these methods are largely static and cannot be updated as newly validated cases accumulate. To address these challenges, we propose Memoir, a memory- driven framework for identifying false positives by transform- ing historical FP alerts into reusable semantic memories. It consists of two key modules. First, historical semantic memory construction converts historical FP alerts into structured semantic memories through LLM-guided annotation, pattern clustering, and memory synthesis to capture reusable behavioral patterns. Moreover, memory-driven identification and evolution retrieves relevant memories and performs semantic verification against taxonomy consistency and security invariants before making the final prediction. It then incorporates verified predictions back into the memory repository, allowing the knowledge base to evolve as new cases accumulate. We evaluate Memoir on CWE- Bench-Java to demonstrate its effectiveness in real-world security analysis. Specifically, Memoir achieves an F1-score of 99.43% with a Recall of 98.88% and perfect Precision, consistently outperforming other baselines. Furthermore, an industrial case study on production software systems from a top IT company shows that the learned memory base generalizes effectively across different SAST tools without retraining.

Visual token compression for vision--language models (VLMs) has largely relied on criteria such as attention, redundancy, and uncertainty to maximize average accuracy under a fixed compute budget, implicitly assuming that all errors carry equal cost. However, the consequence of an incorrect prediction on downstream tasks is rarely symmetric: misreading an invoice amount can be far more costly than misclassifying a background color. Motivated by this, we introduce consequence-sensitive visual token compression, which allocates visual computation across requests according to their potential error costs. Our method follows a calibrate-then-allocate procedure, estimating consequence-specific error-budget curves offline and applying the calibrated token budgets online using consequence signals available from question or task information. On a controlled within-task benchmark, high- and low-consequence questions are drawn from the same document images, so content alone cannot reveal which questions are costly to get wrong. In this setting, our method reduces high-stakes errors from 0.300 to 0.133 under the same total token budget, whereas a content-driven allocator performs no better than uniform allocation. Measuring how error rates change with token budget across different cost ratios, we derive an allocation frontier: uniform allocation is optimal when errors are equally costly, and token transfer toward high-consequence questions becomes increasingly beneficial as the cost gap grows. This allocation principle generalizes well across three dense vision-language benchmarks, two budget realization mechanisms (token deletion and resolution reallocation), two VLM architectures, and multiple token selection strategies. On a realistic mixed workload, consequence-sensitive allocation reduces cost-weighted error by 38% while achieving approximately 21% lower latency than full-resolution inference.

To address the degradation of bearing fault diagnosis accuracy under strong noise, this paper proposes a time-frequency dual-domain multi-scale convolutional neural network. The time-domain branch employs three parallel convolutional kernels to capture multi-scale impulse features, while the frequency-domain branch applies the Fast Fourier Transform to extract noise-robust spectral structure information. Features from both branches are fused for fault classification, yielding a compact model of 110,122 parameters. Experiments on the CWRU bearing dataset across seven signal-to-noise ratio levels demonstrate that the proposed method achieves 99.75% accuracy under clean conditions and maintains 92.50% at -4 dB SNR, representing a 7.25 percentage-point improvement over the single-domain baseline with monotonically increasing gains under stronger noise. Ablation experiments validate the independent performance contributions of the time-domain multi-scale branch and the frequency-domain branch. Comparative experiments against WDCNN, DRSN-CW, MCNN, and 1D-LeNet confirm the superiority of the proposed method under strong noise conditions.

Agent skills are increasingly used to equip large language model (LLM) agents with reusable procedural knowledge. Although recent work has substantially improved skill retrieval due to the increasing skill libraries, retrieving a plausible skill bundle does not guarantee that executing it is worthwhile. Since every skill-conditioned rollout is computationally expensive, deciding whether a retrieved bundle should be executed has become an increasingly important challenge. To this end, we introduce the Reward-Aware Dynamic Execution Gate (RADEG), a lightweight, retriever-agnostic decision layer between skill retrieval and agent execution. RADEG learns a low-cost surrogate model that predicts the execution utility of a query--bundle pair before the expensive rollout is launched. To obtain informative supervision while controlling for task difficulty, we locally perturb each retrieved bundle by deleting, adding, or replacing one skill, producing matched same-query rollouts that isolate the effect of bundle composition on verifier reward. During deployment, RADEG updates only a warm-started logistic head as new verifier feedback becomes available, enabling inexpensive adaptation of the execute/skip boundary without retraining either the retriever or the agent. Under a query-level held-out evaluation on 288 collected rollouts, RADEG substantially reduces unnecessary agent executions while preserving a large fraction of the downstream verifier reward. It consistently outperforms relevance-based and random gating across different execution budgets, demonstrating that execution-aware surrogate modeling provides a practical and cost-effective complement to skill retrieval.

Reliable uncertainty representation is essential for deploying autonomous systems that interact with their environment, as robots must reason about how uncertainty arising from both stochasticity and model mismatch is impacted by contacts with obstacles (e.g., when navigating through a cluttered environment or inserting a part into an assembly). We propose Calibrated Particle-sets for Trans-dimensional Uncertainty Representation (CaPTURe), a geometry-aware, conformal prediction-based algorithm that generates probabilistically valid prediction regions of the unknown future system configuration using particle-based models of arbitrary fidelity. While calibrated uncertainty predictions are essential for safe and efficient planning, analytical or learned motion models are often inaccurate - due to limited data, simplifying assumptions, unmodeled effects, etc. - which can lead to unsafe executions or task failure. Additionally, when a robot contacts an obstacle, the distribution of its future configurations can become multimodal or disjoint, or lie along manifolds of lower intrinsic dimension than the space of possible robot configurations. Our method uses a calibration dataset of system transitions to locally calibrate motion uncertainty estimates, constructing regions guaranteed to contain the future robot configuration at a user-set probability. Our calibration procedure captures how motion uncertainty varies between contact-rich and contactless motions, leading to sufficient coverage in both cases. We evaluate our method on two simulated planning tasks: controlling a marble around a labyrinth and performing tight-tolerance peg-in-hole insertion with a manipulator. Compared to relevant baselines, CaPTURe achieves the user-specified coverage requirement both in and out of contact and achieves up to a 30% absolute improvement in task success rate over the best baseline.

Aligning large language models (LLMs) with human privacy preferences requires capturing individuals' disclosure boundaries beyond general privacy norms. However, a gap remains in eliciting such nuanced preferences to evaluate alignment in realistic settings. We introduce CIDER, a dataset of 14,850 human annotations from 169 users, forming 1,650 contextual disclosure boundary sets across 60 interpersonal communication scenarios involving information sharing that violates privacy norms. Each boundary represents a real user's disclosure decisions over 9 sharing variants in a scenario, for a given communication role and AI-mediated condition. We formulate a task in which models predict a user's disclosure decision from historical boundaries, with varying levels of contextual information. Across 12 open and proprietary models, in-context personalization improves prediction accuracy by up to 11.41 percentage points using only 6 historical examples. Larger models such as GPT-5.4 (with medium reasoning effort) and Claude Sonnet 4.6 are better at leveraging semantic context to understand user-specific, context-dependent disclosure preferences for more accurate predictions, while smaller models tend to rely on structured heuristics based on disclosure granularity and identifiability. Personalization generally improves prediction accuracy, but the improvement is often accompanied by imbalanced shifts in false-positive and false-negative rates across models, with only Claude Sonnet 4.6 achieving balanced improvements in both. Our findings reveal both the promise and limitations of inference-time personalization for privacy preference modeling and position CIDER as a resource for advancing personalized privacy alignment.

Tabular data presents unique challenges for deep learning due to its heterogeneous nature, where numeric features exhibit diverse distributions, scales, and statistical properties. Although recent advances have improved how models learn from tabular data, how numeric data are transformed into model-friendly representations remains comparatively underexplored. We introduce the stretch transformation framework, which formulates numeric feature preprocessing as an optimization problem to make the target function smoother and thus more learnable. Our framework has two variants: (1) unsupervised stretch, which uniformly redistributes feature density via minimax optimization, and (2) supervised stretch, which optimizes target-aware numeric feature transformations from the perspective of target-function smoothness by minimizing the target function's Dirichlet energy in the transformed space. Our theoretical analysis further connects this framework to several popular transformations: unsupervised stretch is closely related to Piecewise Linear Encoding through a shared piecewise-linear geometry and approaches the empirical CDF transformation as the number of bins grows, while supervised stretch becomes closely related to target encoding in the fine-binning limit. Comprehensive experiments on 38 datasets from the TALENT benchmark demonstrate that supervised stretch consistently outperforms all baselines. These results show that explicitly optimizing for target function smoothness is a powerful and underexplored strategy for tabular deep learning.

Query-Key Normalization (QK-Norm) improves the training stability and quality of modern Large Language Models (LLMs). However, under Tensor Parallelism (TP), layerwise QK-Norm introduces additional cross-GPU communication because the normalization factor depends on the full hidden vector. We present SwiftQK, a multi-GPU RMSNorm kernel that exchanges only scalar normalization statistics and overlaps the remaining Peer-to-Peer reduction with independent element-wise computation in a deadlock-safe persistent kernel. Evaluations on recent LLMs show that SwiftQK reduces QK-Norm latency by 81.4--93.9% relative to the standard TP QK-Norm using full-vector All-Gather. In end-to-end serving, SwiftQK reduces TPOT on average by 29.5% over the All-Gather-based baseline and by 14.3% over an optimized scalar-aggregation implementation.

Large audio-language models (LALMs) have demonstrated strong capabilities in understanding diverse audio inputs. This diversity includes low-frequency signals that are inaudible to humans but can still enter the model and influence its generation. However, the practical impact of such low-frequency inputs on LALMs remains largely unexplored. In this paper, we propose Intermittent Low-Frequency Lockout (ILL), an inaudible red teaming method that evaluates this risk using a universal waveform template in a black box setting. ILL uses Sentence Attention Scale Estimation to determine active intervals and Frequency Confusion Transfer to construct a low-frequency waveform with continuous phase from corpus spectral variation. To mitigate this risk, we propose Distributional Requery Guard (DRG) to detect low-frequency distribution shifts and conditionally request a second recording for semantic recovery. Across six LALMs and multiple audio understanding tasks, ILL reduces accuracy by up to 67 percentage points while receiving a mean human audibility rating of 1.33, close to 1.17 for clean audio; DRG raises mean attacked accuracy from 28.5\% to 46.1\% after clean reacquisition. These findings identify a previously overlooked safety risk for LALMs and provide a foundation for future research on robust audio understanding.

Large language models (LLMs) are increasingly expected to follow long lists of constraints in complex instructions, and synthesizing instructions from a reference document (i.e., back-translation) is a widely used method to measure/enhance LLMs' ability to follow complex instructions. However, this method introduces a critical loophole: the constraint synthesis model copies text from the reference as a very specific constraint and the evaluated LLM trivially satisfies the constraint by copying its text in the response. To address these issues, we propose UNSPECIFIC, a novel framework that synthesizes constraints common to two similar reference articles to reduce copy-pasting, selectively hardens only trivially satisfied constraints to balance difficulty and naturalness, and evaluates satisfaction on both the generated article and its summary to penalize superficial instruction following. Consequently, we built the UNSPECIFIC benchmark on news, story, and blog domains to analyze the copy-pasting behavior of LLMs. Our results show that our synthesized constraints are not only more challenging (e.g., the satisfaction rate of GPT-5 Mini drops from 90% to 78%) and natural (LLM win-rate gap improves by 30%) from a human perspective but also mitigate the copy-pasting. We also find that a large portion of constraints are satisfied superficially (i.e., not satisfied in the core narrative of the article). The code and datasets are released at https://github.com/JeetDSharma/UNSPECIFIC.

Production AI agents fail when their context sources -- system prompts, knowledge bases, tool descriptions, and procedural skills -- contain errors or gaps. Current maintenance relies on manual log review and ad-hoc debugging, creating a scalability bottleneck as interaction volume grows. We present TRACE (TRajectory Attribution for Automated Context Engineering), an automated feedback loop that mines historical agent trajectories to diagnose and remediate context failures. Our key insight is that trajectories are rich with implicit dissatisfaction signals -- user corrections, rephrasing, abandonment cues -- that reveal precisely where context sources failed, without explicit feedback collection. Unlike model fine-tuning, TRACE operates on the context layer, enabling rapid iteration without retraining. We make four contributions: (1) a trajectory mining framework that systematically extracts diagnostic information from historical agent executions; (2) multi-component causal attribution that extends textual gradients from monolithic prompt optimization to heterogeneous context sources (skills, knowledge bases, tools, prompts); (3) exploratory verification, where agents actively read context sources to distinguish content gaps requiring CREATE from stale content requiring UPDATE, achieving 96% operation accuracy; and (4) a reusable simulation methodology and verifiable benchmark addressing the absence of open datasets for context debugging, with a six-category fault taxonomy, ground truth annotations, and a cross-layer verification protocol. On 60 dissatisfaction traces spanning three complexity tiers (up to 16 execution nodes), TRACE achieves 72.7% root cause attribution and 82% end-to-end fix effectiveness, showing that over 80% of context-layer failures can be automatically diagnosed and remediated by mining historical trajectories, an overlooked resource in production systems.

Treatment planning in precision oncology requires synthesizing heterogeneous patient information with rapidly evolving clinical guidelines to ensure guideline-concordant care. While large language models (LLMs) show promise in many diagnostic tasks, their adoption for high-stakes treatment planning is hindered by complex reasoning, adherence to timely clinical guidelines, and safety concerns. In this study, we present GatorOnco, an agentic LLM for colorectal cancer (CRC) treatment planning. GatorOnco is developed using a total of 282 billion tokens of biomedical text, including healthcare system-scale clinical text comprising 166 billion tokens from UF Health. We implemented a domain-adaptation method that integrates pre-training, model merging, a two-stage post-training approach, and agent-based reinforcement learning. An agentic retrieval-augmented generation (RAG) approach dynamically integrates time-sensitive clinical guidelines into the reasoning process. In a blind, randomized clinical evaluation conducted by five UF Health oncologists, GatorOnco significantly outperformed open-source LLMs (P < 0.01) and achieved expert-level performance comparable to UF Health oncologists. Compared with expert oncologists, GatorOnco received significantly higher ratings for readability (4.46 vs. 4.19, P < 0.01) and completeness (3.91 vs. 3.52, P < 0.01), while showing statistically comparable performance in correctness (4.09 vs. 4.11, P = 0.921), currency (4.04 vs. 3.98, P = 0.478), and safety (4.22 vs. 4.22, P = 0.999). These findings demonstrate that integrating agentic reasoning with large-scale domain adaptation can help bridge the gap for generative AI in high-stakes cancer treatment planning.

Recent work on protecting privacy during user-LLM interactions often focuses on direct, explicit identifiers: the personally-identifiable information (PII) captured by standard detectors. One such approach is Privacy-Conscious Delegation (PCD), where a local LLM acts as an intermediary. However, privacy risk does not stem solely from explicit identifiers but also PII-free self-disclosures, leaving users identifiable through combinations of quasi-identifying traits. We investigate a probabilistic variant of PCD, where we augment its objectives with an LLM-driven probabilistic estimation of k-anonymity. To facilitate this, we first create the PUPA-SD dataset, which contains naturalistic user queries with self-disclosure. Our preliminary results indicate that optimizing PAPILLON on PUPA-SD improves quality on unseen conversations across a variety of local models and produces the best privacy-utility balance for Llama-3.2-3B, while smaller models struggle to jointly optimize quality and privacy. We propose k-anonymity as a useful auxiliary metric for tackling PCD.

While learned robotic policies hold promise for advancing generalizable manipulation, their practical deployment is often hindered by suboptimal execution speeds. Imitation learning policies are inherently limited by hardware constraints and the speed of the operator during data collection. In addition, there are no established methods for accelerating policies learned via imitation, and the empirical relationship between execution speed and task success remains underexplored. To address these issues, we introduce SpeedTuning, a reinforcement learning framework specifically designed to enhance the speed of manipulation policies. SpeedTuning learns to predict the optimal execution speed for actions, thereby complementing a base policy without necessitating additional data collection. We provide empirical evidence that SpeedTuning achieves substantial improvements in execution speed, exceeding 2.4x speed-up, while preserving an adequate success rate compared to both the original task policy and straightforward speed-up methods such as linear interpolation at a fixed speed. We evaluate our approach across a diverse set of dynamic and precise tasks, including pouring, throwing, and picking, demonstrating its effectiveness and robustness in enhancing real-world robotic manipulation.

Image super-resolution (SR) with large generative models has recently achieved remarkable perceptual quality, yet maintaining fidelity to the LR observation remains challenging. In particular, we observe that diffusion transformers (DiTs) built on latent representations suffer from a critical limitation: the compression bottleneck of the VAE weakens fine-grained spatial information, leading to hallucinated details that are weakly grounded in the input image. In this work, we revisit generative SR from a representation perspective and propose a pixel-grounded super-resolution (PGSR) framework that preserves LR-observed pixel evidence before VAE compression and reuses it throughout restoration. Instead of relying solely on the compressed latent condition, PGSR extracts pre-VAE pixel evidence from the upsampled LR image and reuses it at two stages. First, Condition-Side Trajectory Guidance fuses LR-derived pixel evidence with the latent LR condition to guide the latent restoration trajectory. Second, Decoder-Side Pixel Grounding injects multi-scale pixel features into the frozen VAE decoder to ground the final rendering with LR-observed cues. To efficiently adapt large pretrained DiT models, we keep the latent autoencoder and main flow-matching backbone frozen, and train only lightweight restoration modules. We further study an efficient local-window attention variant for improved high-resolution efficiency and scalability. Extensive experiments demonstrate that PGSR improves the realism--fidelity trade-off and produces more faithful, visually convincing results than existing latent generative SR approaches.

Allocating limited computation among concurrent learning tasks is difficult when each task must reach a target loss before a deadline but its required training effort is unknown. Existing approaches combine online loss prediction with adaptive resource allocation, yet commonly treat computation as continuously divisible throughput. We instead study a practical setting in which tasks arrive over time and computation is provided by discrete nodes. This setting introduces both uncertain demand and constrained sequential decisions. We propose MARA, which predicts future loss trajectories with conditional flow matching and coordinates compute nodes through a cooperative multi-agent autoregressive policy. A potential-based progress reward supplies intermediate training feedback while preserving the undiscounted task-completion objective. Across in-distribution, reinforcement-learning, and vision workloads, flow matching reduces remaining-resource prediction error relative to weighted least squares. At the scheduler's training load, MARA completes 63.46% of tasks on average, 8.54 percentage points above strong baseline Learning with Adaptive Resource Allocation (LARA), and remains ahead under unseen heavier workloads.

LLM agents are increasingly deployed in multi-agent social settings where they must cooperate, negotiate, and adapt to other agents. Measuring and improving these social skills is hard because, unlike math or logic, social interaction offers no objective ground truth: evaluations fall back on LLM judges, which are costly, subjective, and noisy, and models get no reliable signal to learn from. To address both, we first introduce Social Gym, an environment of 21 multi-agent social games (e.g., Werewolves, Resistance, Spyfall) whose rule-decided outcomes make agent performance verifiable and objective, with an Elo tournament that produces a cross-game leaderboard. Benchmarking experiments show that while GPT-5-mini tops the leaderboard, no model excels at all games uniformly or in all game roles, pointing to limitations of social reasoning. Motivated by this, we additionally propose SPaRTan (Self-Play and Reflect-Transfer), a training-free self-improvement loop: a model plays a game, reflects on its trajectories and their outcomes to produce a transferable playbook, and applies that playbook in subsequent games. Our results show that SPaRTan playbooks help GPT-5-mini agents level their performance on weaker roles, but largely do not improve Qwen3-32B's performance. Together, Social Gym and SPaRTan offer a reproducible, verifiable foundation for measuring and improving LLM social reasoning without weight updates.

In this project, we delved into the pervasive challenge of bias detection within the text content. More specifically, our focus lies on the identification of subjective bias, a type of bias that introduces improper attitudes or portrays a statement at odds with the actual truth. The subjective bias can jeopardize the authenticity and reliability of texts, leading to misconceptions and potential social tensions, especially when expressed through offensive language. Following prior work [1], we tackled with three different types of subjective biases in text: (1) framing bias with the use of one-sided words or phrases containing a particular point of view; (2) epistemological bias which includes subtle linguistic features that can affect the believability of the texts; (3) demographic bias with word/phrase usage under presuppositions of a particular demographic factor (i.e., gender or religion). In terms of the data we utilize, the input consists of texts that may harbor subjective biases. The output is a classification or annotation that reveals the presence or absence of such biases within the provided content. More specifically, we detected three different types of multi-span biases in corpus WIKIBIAS [2] with more than 4,000 sentence pairs from Wikipedia edits. The data is labelled by bias type for span pairs with the following categories: (1) framing bias, (2) epistemological bias, (3) demographic bias, and (4) no bias. The project codes are released at https://github.com/HoningJade/LLM-Bias-Type-Classification.

Temporal decisions in language-model systems often depend on both symbolic task state and elapsed wall-clock time, such as cache expiration, job completion, quota resets, deadlines, or stale sessions. We study whether elapsed time can be supplied as a non-token, system-side scalar and composed with visible symbolic state by a frozen-backbone language model. We introduce ChronoState, a compositional temporal-state benchmark in which symbolic state appears in the prompt, elapsed seconds tau are supplied through a hidden chronometric-injection channel, and the model selects a forced-choice temporal action. Here, "hidden" means hidden from the user-visible token sequence, not from model computation. Using Qwen2.5-3B-Instruct as a frozen bf16 backbone with a 31-dimensional sinusoidal-plus-log time encoding, gated FiLM residual modulation, and a rank-8 LoRA action surface, hidden-time CI reaches 0.9305 +/- 0.0134 accuracy and 0.9410 +/- 0.0103 balanced accuracy. No-time and shuffled-time controls fall to 0.5511 +/- 0.0042 and 0.3323 +/- 0.0097, respectively, with high shuffled-time wrong-state consistency supporting causal dependence on the injected scalar within the trained distribution. Generalization remains strong for held-out templates, durations, and multi-constraint compositions, but held-out quota-family transfer is weak at 0.5065 +/- 0.0559, while a fair prompt+LoRA timestamp baseline reaches 0.9893 +/- 0.0052. Thus, ChronoState supports a narrow conclusion: hidden elapsed time can be composed with symbolic task state under direct supervision, but does not establish autonomous time tracking, broad unseen-family abstraction, or superiority over prompt-injected timestamps.

Aligning Large Language Models (LLMs) for open-ended tasks is challenging because responses must satisfy multidimensional criteria without following a single correct generation trajectory. Existing rubric-based reinforcement learning (RL) methods compress fine-grained criterion-level feedback into scalar rewards, making persistent capability gaps difficult to target under limited on-policy exploration. We propose $\textbf{RISE-RL}$ (Rubric-Informed Selective Exploration), which uses repeatedly missed rubric criteria to elicit privileged trajectories that are difficult to discover through unguided exploration alone. RISE-RL retains only trajectories whose complete-rubric reward exceeds the mean reward of natural rollouts, and then re-evaluates them under the original prompt to emphasize behaviors that remain weakly supported by the natural policy. The resulting guidance signal is optimized through a separate auxiliary objective and removed once its additional benefit diminishes. Experiments with 4B and 14B models across writing, chat, health, and science show that RISE-RL achieves the highest mean score on every evaluated benchmark under guidance-free evaluation. Compared with standard Rubric-RL, it improves the average score by 1.3 points at the 4B scale and $\textbf{3.3 points at the 14B scale}$, including a $\textbf{6.0-point}$ gain on CreativeWriting-V3. It also improves creative-writing diversity and yields gains on objectively scored medical and scientific benchmarks. These results indicate that selective internalization through reward filtering and policy support shaping is effective for open-ended reinforcement learning.

The localized depiction of perceptual quality has long been a crucial, yet underexplored, challenge in image quality assessment (IQA). Existing approaches based on large multimodal models (LMMs) predominantly rely on text-driven supervised fine-tuning (SFT). However, this training paradigm exhibits notable limitations in detection accuracy. Moreover, synthetically distorted images, which are often used as the primary training data source, show a significant generalization gap when deployed in real-world scenarios; thus, the \textbf{synthetic-to-authentic (\textit{S2A})} problem represents a critical challenge. Motivated by these issues, we propose \textbf{\textit{VIGIL}}, which leverages the LMM architecture for precise visual distortion detection. From a candidate pool of over 1000K samples, we construct the \textbf{\textit{VIGIL-140K}} training set, which consists of over 140K distorted images. These images are obtained through rigorous quality filtering and carefully crafted distortion injection, covering 8 major synthetic distortion categories. Our model leverages different layers of the large language model (LLM) decoder, treating them as \textit{multiple detectors} that perform synchronous distortion detection using multi-level features. Additionally, we retain distortion cues from predictions assigned to the non-distortion class, which helps mitigate the ambiguous foreground-background (\textit{FG-BG}) separation commonly encountered in the \textit{S2A} problem. After post-processing, our model consistently outperforms strong baselines on both in-domain synthetic distortion detection and \textit{S2A} tasks.

Web navigation agents are capable of addressing various types of tasks on different websites. Current baselines on web navigation are either unimodal or lack strong reasoning abilities given multimodal inputs. Focusing on the WebShop benchmark, a real-world website simulation, we explore the alignment of text and images, as well as multimodal reasoning and planning abilities, to enhance the performance of web navigation agents. We propose three innovative multimodal enhancements: Multimodal Enhanced LLM for Online Navigation (MELLON), VQAgent, and Multimodal Ranker. MELLON demonstrates a significant improvement in task completion accuracy, with a 9.26% increase after just one epoch of training. Our findings suggest the necessity of further exploration into multimodal approaches, with a focus on more extensive training and alignment strategies to enhance the effectiveness of web navigation agents.

We introduce Motif 3, a decoder-only Mixture-of-Experts language model with 314 billion total parameters and 13.2 billion activated per token. Each sparse MoE layer contains 384 routed experts, with eight selected per token. This fine-grained sparsity provides substantial expert capacity while limiting computation. Motif 3 is built around Grouped Differential Latent Attention (GDLA), which integrates grouped differential attention with the compressed key-value representation of Multi-head Latent Attention. The architecture further incorporates modified manifold-constrained hyper-connections, Expert Specific PolyNorm activations, and multi-token prediction to improve optimization stability, expert specialization, and inference efficiency. We pretrain Motif 3 on approximately 12.5 trillion tokens spanning web documents, STEM, code, mathematics, multilingual content, and domain-specialized corpora. Expert-balancing and numerical-stabilization techniques support stable training at scale, while selective MXFP8 computation and communication, memory-efficient fused kernels, and window-aware context parallelism enable training with context lengths up to 256K tokens. Our post-training pipeline combines general supervised fine-tuning, six specialist teachers trained with reinforcement learning, a software-engineering teacher trained with supervised fine-tuning, and Multi-teacher On-Policy Distillation. The resulting unified model consolidates complementary capabilities in reasoning, coding, tool use, professional work, long-context understanding, calibrated abstention, and instruction following. Across a broad evaluation suite, Motif 3 demonstrates competitive performance against leading open weight models, including strong results on long-horizon agentic tasks, mathematical reasoning, scientific knowledge, and hallucination-sensitive evaluation.

Probabilistic Circuits (PCs) are tractable generative models whose internal nodes encode a hierarchy of probabilistic sum- maries over different variable scopes. Existing PC-based out- of-distribution (OOD) detection methods ignore this hierar- chy, reducing the entire circuit to the scalar likelihood (or its uncertainty) computed at the root. We introduce Hierar- chical Likelihood Vector (HLV), a representation whose en- tries are the likelihoods associated with selected PC nodes and define the Hierarchical Likelihood Distance (HLD), a PC-induced pseudo-metric that compares the probability dis- tributions through the expectations of their HLVs. We show that HLD is an integral probability metric over a function class naturally induced by the PC and develop a principled goodness-of-fit hypothesis test for unsupervised OOD detec- tion. Unlike existing approaches, the trained PC alone serves as the representation of the in-distribution: no held-out in- distribution data are required at deployment. We further show that the quantities required by the hypothesis test can be com- puted exactly, directly from the trained circuit, yielding an ap- proximate analytic decision threshold. Experiments on tabular and MNIST datasets demonstrate that exploiting the hierarchi- cal probabilistic summaries encoded through the PC improve OOD detection over root-likelihood, uncertainty-, typicality- and kernel-based baselines, while naturally localizing distri- bution shifts to the PC nodes responsible for the shift.

AI video generation has advanced rapidly and entered widespread commercial use. As a result, quality differences among videos produced by state-of-the-art AI video generation models~(AIVGMs) have become increasingly difficult to discern using conventional evaluation criteria, such as visual fidelity and semantic instruction following. Meanwhile, human evaluation now requires more expertise and sustained attention, substantially increasing annotation costs. This calls for automated evaluation that can reliably distinguish fine-grained differences among advanced AIVGMs with minimal human intervention. To address this challenge, we present RAVEN-Eval, a rubric-guided automated evaluation framework for AIVGMs, built primarily on the LMM-as-a-judge paradigm. Through an automatic task curation and quality-filtering pipeline, RAVEN-Eval curates 150 text-to-video~(T2V) tasks and 100 image-to-video~(I2V) tasks, and systematically collects more than 4,500 AIGVs. At its core, RAVEN-Eval adopts rubric-guided automated LMM preference judgement, in which LMM judges conduct pairwise comparisons according to task-specific rubrics. It further introduces an anchor-based model insertion approach to reduce the evaluation cost of incorporating new models. Finally, we evaluate 20 high-performance AIVGMs, as well as the judging capabilities of 13 LMM judges, and establish the RAVEN-Eval Leaderboards. Overall, RAVEN-Eval paves a scalable path for automatic and trustworthy evaluation of rapidly evolving AIVGMs.

User feedback offers natural supervision for persistent LLM improvement, but a single message may support multiple behavioral changes with different scopes of generalization. We introduce SLIFT, a selective self-learning framework built on a task-relative view of user feedback. SLIFT decomposes each feedback message into atomic components and interprets each component relative to the original task as Fix, Spec, or Null: requirements for task validity, compatible condition-specific refinements, or content with no reliable positive update direction. To incorporate each change at the appropriate scope, SLIFT trains two complementary LoRA adapters on a shared frozen backbone: a Generalist that consolidates Fix requirements into default behavior through feedback-conditioned self-distillation, and a Specialist that observes only the task and Generalist response to supply residual guidance for applicable, unmet Spec refinements. Null components induce no positive update. Across backbones, SLIFT achieves strong performance on both MemoryBench and WildFB, with targeted analyses further examining its underlying mechanisms. We release our code at https://anonymous.4open.science/r/SLIFT.

Large language models (LLMs) have demonstrated strong performance across a wide range of legal tasks. In legal practice, time is a critical concept that governs the validity of statutes, the progression of legal cases, and the enforcement of procedural deadlines. However, legal temporal capabilities remain underexplored in existing legal AI benchmarks. To address this gap, we propose LexKairos, a comprehensive benchmark for evaluating the temporal capabilities of LLMs in the Chinese legal context across three dimensions: statutory temporal knowledge, case temporal modeling, and statute-case temporal reasoning. LexKairos comprises nine sub-tasks drawn from real-world Chinese judicial cases and statutes. We conduct systematic evaluations of eight LLMs under multiple inference settings, including vanilla, Chain-of-Thought (CoT), and thinking modes. Our results show that Gemini-3-Flash achieves the strongest overall performance, yet even the best-performing model exhibits notable limitations on tasks demanding precise time-sensitive statutory metadata recall or complex reasoning in time limits, indicating that legal temporal knowledge and reasoning remain open challenges for current LLMs. Data and code are available at https://github.com/thunlp/LexKairos.

Scanning probe microscopy provides nanoscale access to structural, electrical, electromechanical, magnetic, and mechanical properties of materials. Its increasing use for wafer-scale characterization and combinatorial materials exploration creates a need to distribute measurements efficiently across large spatial domains. This is particularly important when available modalities differ in acquisition time and potential for tip and sample damage, making exhaustive multimodal mapping over spatial grids impractical. Here, we demonstrate multitask scanning probe microscopy, a live, closed-loop workflow in which a multitask Gaussian process learns spatial and cross-modal relationships and autonomously selects both the next measurement location and the next experimental protocol. The approach is implemented on an automated large-sample atomic force microscope and demonstrated on a composition-spread AlScN wafer using tapping-mode and Dual AC Resonance Tracking (DART) measurements. Paired initial measurements establish the relation between the tasks, after which noncoincident measurements are used to update both response landscapes. The resulting workflow extends active learning in scanning probe microscopy from spatial sampling to autonomous allocation of measurement modalities and provides a basis for combining rapid, weakly perturbative imaging with slower contact, electrical, electromechanical, magnetic, or spectroscopic measurements.

Semantic segmentation models are trained and evaluated against human-drawn masks, yet remote-sensing annotations are often coarse, incomplete, or misaligned; high overlap scores may then reflect agreement with imperfect labels rather than faithfulness to the image, creating an evaluation paradox. We introduce Contrastive Mask Fidelity (CMF), a training-free, reference-free metric that scores competing class masks directly against image evidence. CMF composites keep and erase counterfactual views of each mask and asks a frozen vision-language judge whether class evidence is concentrated inside the mask and absent outside. We validate CMF on controlled mask corruptions, then audit 10,731 image-class pairs across ten remote-sensing benchmarks using candidate masks from Seg-Probe, a training-free open-vocabulary probe built on SegEarth-OV3 that outperforms prior baselines on nine of ten datasets. The audit reveals systematic, class-dependent annotation distortion: man-made classes such as buildings, roads, and cars favor the candidate mask on 62-85% of pairs, whereas ambiguous land cover more often favors human annotations. On a blinded three-annotator consensus, CMF matches expert judgment on 81% of pairs, exceeding keep-only scoring, model confidence, and a trained label-quality baseline. Finally, conservative class-wise arbitration yields supervision that improves cross-domain transfer over raw annotations and matched replacement controls, positioning CMF as a scalable tool for auditing ground truth rather than presuming it infallible.

Millions of chemical structures appear in patents and papers only as drawings, and using that information at scale requires reading the drawings. OCSR appears nearly solved on synthetic images yet remains difficult on real documents: the starting recognizer, Qwen2.5-VL-7B, exceeds 91% accuracy on synthetic renders but falls below 16% on three real-world benchmarks (ACS, CLEF-IP, USPTO). To identify the main source of improvement, 21 recognizers were fine-tuned on mixtures of synthetically rendered structures and labeled real depictions from patents, journal figures, and hand-drawn collections, varying the vision language model (VLM) base, the fraction of real training data, and the vision-tower adaptation strategy. Labeled real training images make the largest difference. For Qwen2.5-VL, ACS exact match rises from 0.15 with no real data to 0.37 at 9.5% and 0.46 at 50.2%; a controlled experiment across three base models reproduces the trend. A vision-tower LoRA, in contrast, does nothing for Qwen (+0.00, paired p=1.00), substantially helps InternVL3-8B (+22.8 to +34.6 pt), and modestly helps GLM-4.1V-9B (+1.0 to +9.6 pt), so its value depends on the base model. The best configuration reaches 0.96 exact match on clean renders and 0.49, 0.65, 0.84, and 0.76 on ACS, CLEF-IP, UOB, and USPTO, respectively. Gaps between base models are largest without real data (0.21), shrink to 0.06 at 70% real data, and reorder the ranking; base model and real-data mixture must therefore be selected together. Small-scale experiments on handwritten image-to-LaTeX recognition and chart-to-table conversion show that base-model rankings also vary beyond chemistry. More generally, model and adaptation choices for visual structure recognition should be evaluated on the target task.

Quantitative estimation of protein-ligand binding affinity from three-dimensional complex structures is a fundamental task in structure-based computational chemistry and molecular modeling. Reliable prediction remains challenging because available structure-affinity data are limited, experimentally heterogeneous, conformation-dependent, and sensitive to dataset partitioning. RAVEN (Randomized Atomistic Views with Ensemble Neural Reservoirs) utilizes a multihead reservoir of independently initialized and fully frozen atomistic graph encoders to generate diverse structural projections without end-to-end optimization of the graph representation. These projections are integrated with a deterministic physicochemical interaction fingerprint and processed by heterogeneous supervised readers, including neural and tree-based regressors, whose outputs are combined through validation-based nonnegative fusion. The random reservoir expands structural feature coverage across independent encoder realizations, whereas the explicit physicochemical descriptors and heterogeneous readers contribute complementary information and distinct inductive biases. Evaluation on a similarity-isolated PDBbind 2020R1 split reconstructed using GEMS similarity resources, together with the protected CASF-2016 subset, demonstrated strong predictive performance. The results indicate that frozen multi-view graph representations, explicit physicochemical statistics, and heterogeneous model fusion provide a robust and flexible framework for protein-ligand binding-affinity prediction.

Large Language Models (LLMs) have driven rapid progress in autonomous agents, yet standard evaluations remain confined to static task solving. An emerging frontier is harness evolution---the agent's capacity to autonomously optimize its own operating harness. However, systematically benchmarking this capability remains challenging, as existing evaluations fail to isolate harness improvements from base model strength, prevent task-specific overfitting, or capture long-horizon iterative research. To address these challenges, we introduce Evo-Bench, the first benchmark designed to evaluate models' intrinsic harness-evolving capabilities across Search, Office, and General agent domains. To rigorously isolate this capability, Evo-Bench employs a novel harness-guided construction framework: it leverages auxiliary-task evolution to identify tasks genuinely sensitive to framework improvements, followed by sensitivity-aware stratified splitting to ensure robust cross-suite generalization. Extensive evaluations across nine frontier and open-weight models reveal that top models achieve massive absolute gains reaching 16.6 points, closely approaching state-of-the-art human-engineered baselines. Crucially, while autonomous evolution outpeforms artificial harness in General tasks and excels in Search tasks, it struggles in Office tasks that demand highly specific processing workflows. Furthermore, our analysis exposes critical temporal anomalies like early saturation, while demonstrating that the synthesized harnesses act as highly transferable reasoning structures, consistently boosting diverse policy models.

Uncovering the internal mechanisms underlying the safety capabilities of large language models (LLMs) is crucial for developing trustworthy artificial intelligence. Currently, mechanistic interpretability studies on multilingual safety are largely confined to local components, such as isolated neurons. However, this static and fragmented perspective overlooks the synergy among components and fails to elucidate how safety signals dynamically propagate within the model to drive safety decisions ultimately. In this work, we move beyond isolated neurons to identify and target the cross-layer functional pathways formed during safety signal propagation, thereby uncovering the mechanisms driving the cross-lingual safety gap. Specifically, we first identify monolingual safety pathways and validate their impact on refusing harmful requests. Subsequent cross-lingual analyses reveal a sparse subset of cross-lingual shared safety pathways, confirming that this intersection acts as the internal bridge transferring safety capabilities from high-resource (HR) languages to non-high-resource (NHR) languages. Building on these mechanistic findings, we propose a pathways-targeted alignment method based on the cross-lingual shared safety pathways. Experimental results show that updating only a small fraction of pathway parameters significantly improves safety in NHR languages while largely preserving the model's general capabilities.

How a document's arrangement is written down, its notation, is a training variable that no dataset card records. The field has established that text-extraction choices change model behaviour, and has never once measured the notation of what those choices put into the corpus. We define clean-window survival, a deterministic count of how much of a stream still demands the boundary inference, and measure notation on three fronts. What corpora carry: a census of thirteen public corpora, where survival falls to 0.153 in a vision-converted PDF slice against 0.889 in C4; the scarce resource is not unmarked text but long unmarked text; a pre-registered supply test finds what remains institutional, not consumer. Our own pre-registered prediction failed: converters do not fabricate structure on prose, and that null forced the reliability mechanism that survives it. What readers use: across five base models spanning 0.6B to 8.2B and two pipelines, deleting a structural announcement makes the following prose measurably harder to predict, while swapping its notation moves nothing. That zero does not make notation unimportant; it relocates the variable: the operative cue is the announcement, not the sigil. What writers impose: a bounded null. Base models do not impose the marked register above the authored baseline, and handed prose with every announcement deleted they do not put one back, at a rate indistinguishable from zero against an authored reference of zero. We ship the format those measurements imply: the pure frame, paragraphs in authored order, every announcement deleted into a reversible sidecar, mixed against the marked copy over announcement presence rather than notation. Choose format operators by the capability they train, not by the fidelity they preserve, and record extractor identity and survival on data cards.

Transfer learning is particularly useful in settings with limited training data, and within image classification it is common to transfer learn upon massive datasets like ImageNet , CIFAR-100, or COCO . Qualitatively, it seems a transfer learning dataset should have both more classes and more examples per class than the fine tuning dataset; however, a quantitative method to choose the best transfer learning dataset does not currently exist. In this paper, we design TLDChoiceNet, a model to choose the best transfer learning dataset given a fine tuning dataset by predicting the test-set accuracy after fine-tuning. A simple version 1 achieves 0.154 MSE on the test dataset, while a version 2 leveraging an ImageNet pre-trained ResNet50 v2 embedding with per-class information attains a 5X lower MSE of 0.031. We further design two metrics that enable an unsupervised method of choosing an optimal transfer learning dataset: distribution distance (DD), which linearly regresses against fine-tune accuracy with an R2 of 0.89, and average class correlation (ACC), which improves the R2 to 0.97. Our results underscore that a dataset's low-level statistics can explain the transfer learning effect, and that using a pre-trained ImageNet can embed different classes further apart in latent feature space.

Mixed emotions represent a clinically relevant but still underexplored target for automatic emotion recognition. EEG provides millisecond-level access to neural activity, yet most EEG pipelines analyze the signal through a single temporal window, thereby fixing the temporal structure available to the model. This study introduces a multi-scale temporal framework for EEG-based emotion recognition. The EEG waveform is decomposed into windows of one or several durations, processed by a shared attention-based encoder, and integrated through a dynamic fusion module that assigns sample-specific weights across temporal scales. The framework is evaluated under a subject-independent protocol in binary and three-class settings, with the three-class task including the mixed affective category. The best results are 65.22% for the two-class task and 45.43% for the three-class task. Both are obtained with three-scale dynamic-fusion configurations and remain substantially above the full-signal baseline. The best-performing temporal scales differ between the two tasks. Dynamic fusion outperforms concatenation in the highest-scoring two-class configuration and slightly exceeds it in the highest-scoring three-class configuration, although these multi-scale settings require substantially more computation than the full-signal baseline.

Spatiotemporal prediction on graph-structured data is central to traffic forecasting and environmental monitoring, yet decentralized and heterogeneous data complicate both sequence modeling and collaborative training. We propose F$^2$STNet, a federated forecasting framework that combines truncated graph-Fourier features, a lightweight diagonal state-space temporal encoder, graph convolution, and Fairness-aware Federated Aggregation (FFA). The spectral branch exposes graph-frequency structure, while the state-space layer models long temporal dependencies with linear complexity in the sequence length. FFA adjusts the FedAvg prior using client validation losses and an increasing fairness schedule. Experiments on PeMS04, HZMetro, and KnowAir show favorable forecasting accuracy relative to the evaluated baselines; federated experiments on PeMS04 additionally improve worst-client and client-dispersion metrics.

Large Language Models (LLMs) have achieved strong performance in medical question answering and clinical reasoning tasks. However, their reliability under uncertainty remains poorly understood which raises critical concerns for deployment in high-stakes clinical settings. In such environments, incorrect predictions are inherently risky, but confident incorrect predictions can be particularly harmful as they may mislead clinical decision-making. In this paper, we conduct a systematic behavioral analysis of LLMs under clinical information uncertainty. We propose an evaluation framework based on the MedMCQA dataset consisting of two complementary uncertainty settings. First, we introduce linguistic uncertainty cues through prompt modifications to simulate ambiguous clinical contexts. Second, we construct an answer removal setting, wherein the correct option is deliberately excluded mandating the model to recognize insufficient information and abstain. We analyze both model accuracy and confidence behavior using multiple calibration metrics including calibration gap, Expected Calibration Error (ECE), and Unsafe Confident Error Rate (UCER) across 500 medical questions. Our results reveal a consistent failure mode, i.e., although accuracy degrades under increasing uncertainty, model confidence remains misaligned with accuracy. This leads to a substantial increase in unsafe confident errors, indicating that model confidence remains largely insensitive to clinically meaningful information loss. Furthermore, we observe significant variation across models in their ability to abstain when the correct answer is unavailable, with some models persistently producing high confidence hallucinated answers. These findings expose critical limitations in the epistemic reliability of current LLMs and highlight the need for uncertainty aware evaluation methods prior to their deployment in clinical workflows.

We develop a new approach to Personalized Federated Learning across heterogeneous clients using Nonparametric Empirical Bayes (NPEB). Leveraging the asymptotic normality of local parameter estimates obtained from Empirical Risk Minimization or M-estimation, our method formulates these estimates as noisy observations to estimate an unknown shared prior via Nonparametric Maximum Likelihood. A key challenge in applying NPEB in this setting is that existing approaches assume known fixed variances, which is not true in practice. To address this, we introduce a Variance-Aware Nonparametric Empirical Bayes (VANEB) framework that leverages the parameter-dependent asymptotic variance of local M-estimators. A key technical contribution is a generalized Tweedie's formula for this heteroskedastic setting. We then establish non-asymptotic error rates for density estimation in the average squared Hellinger distance and derive an oracle denoising inequality that provides error bounds for our estimator. While our theoretical guarantees are rooted in the asymptotic regime of M-estimators, we empirically explore heuristic extensions of VANEB to modern federated learning settings involving Deep Neural Networks (DNNs). For DNNs, we propose VANEB-head and VANEB-FT, which personalize the last fully connected layer via an NPEB step using an approximate diagonal variance estimator. We show that our method has strong performance on popular vision datasets MNIST and CIFAR-10, using a convolutional neural network architecture.

Large language model-powered coding agents are increasingly used to modify existing code repositories, for example, by adding features or fixing bugs. Yet existing repository-level benchmarks typically evaluate only whether the final patch passes tests. Satisfying a user request requires a long chain of interdependent reasoning and decisions: an agent must recover explicit and implicit requirements, formulate a repository-grounded implementation plan, and translate it into correct code. A pass/fail outcome cannot characterize how an unsuccessful trajectory diverges from the requirements and implementation process needed for a correct patch. To address this gap, we introduce SWE-RPG, a repository-level benchmark that combines executable patch evaluation with validated ground-truth references (GTs) for (1) Requirement Clarification and (2) Implementation Planning. These intermediate GTs support retrospective, GT-aligned diagnosis of complete coding-agent trajectories across clarification, planning, code generation, and artifact submission. SWE-RPG comprises 163 tasks from 31 Python and Java repositories, including 113 bug fixes and 50 feature additions. We evaluate 3 coding agents, including Claude Code, Codex, and OpenCode, with 6 large language model backends, including Claude-Sonnet-5 and GPT-5.6-Terra. Results show that the evaluated popular coding agents still struggle to implement user requests in existing repositories, achieving an average resolved rate of only 31.5% on SWE-RPG. Intermediate-GT diagnosis further identifies implicit requirement recovery as the main bottleneck, accounting for 24.5%--46.0% of agent runs. This result suggests implicit-requirement recovery as a key candidate direction for improving coding agents. The benchmark data and evaluation code are available at https://github.com/Xin-Zhou-smu/SWE-RPG-Bench.

Forward uncertainty propagation in complex physical systems can induce structured covariance across field-valued outputs. For a probabilistic surrogate, the total predictive covariance comprises the covariance of conditional means across input realizations and the average conditional predictive covariance. Probabilistic DeepONet (Prob-DeepONet) provides lightweight uncertainty quantification by predicting pointwise Gaussian means and variances in a single forward pass, but its conditional predictive covariance is restricted to a diagonal form. To represent cross-location conditional dependence without explicitly parameterizing a full high-dimensional covariance matrix, we develop a two-step mean-variance DeepONet (two-step MV-DeepONet) through two principal modifications. First, two-step training is used to decouple output-basis learning from the input-to-coefficient mapping, together with basis orthogonalization and subspace rotation. Second, Gaussian probabilistic modeling is transferred from the high-dimensional physical output space to the low-dimensional rotated coefficient space. Mapping these probabilistic coefficients through the shared basis induces a generally non-diagonal conditional predictive covariance in the physical output space while retaining single-pass inference. A Frobenius-norm error decomposition and corresponding upper bound identify low-rank covariance compressibility, trunk-subspace approximation, finite-sample statistical error, and coefficient-space covariance estimation as the principal factors governing covariance recovery. Numerical experiments on three representative problems governed by partial differential equations (PDEs) and a hypersonic blunt-body aerothermal problem show improved generalization, more structured uncertainty bands, and accurate recovery of off-diagonal correlation patterns compared with Prob-DeepONet.

Large Language Models (LLMs) have achieved impressive performance in natural language-to-code generation; however, their ability to follow structured algorithmic reasoning remains insufficiently understood. We introduce Pseudo2Code, a benchmark designed to systematically evaluate the impact of structured pseudocode on code generation quality and algorithmic faithfulness. The benchmark consists of 300 manually validated real-world programming tasks spanning multiple domains and three difficulty levels (Easy, Medium, and Hard). Each task contains a problem description, structured pseudocode, reference implementation, and executable test suite. To ensure benchmark reliability, we adopt a dual-stage human validation protocol and release fully executable benchmark instances. Beyond the benchmark, we propose the Pseudo2Code Agentic Framework, a multi-stage pipeline that leverages pseudocode as an explicit intermediate reasoning representation for code generation. We evaluate both commercial and open-source language models using a rubric-based evaluation framework that measures correctness, completeness, relevance, clarity, reasoning quality, and pseudocode adherence, complemented by execution-based testing. Experimental results demonstrate that the proposed Pseudo2Code Agentic Pipeline consistently outperforms strong commercial and open-source baselines, achieving an overall score of 4.78 compared to 4.31 for the strongest baseline model. Furthermore, a human evaluation study involving 100 benchmark tasks shows strong agreement between human judgments and automated assessments. Our findings provide empirical evidence that structured pseudocode improves functional correctness, reasoning quality, and algorithmic faithfulness in code generation. We release Pseudo2Code to support future research on structured reasoning, interpretable code generation, and reliable AI-assisted software development.

Cardiologists interpret electrocardiograms by localizing waveform components, measuring rhythm and interval patterns, and translating these structured observations into diagnostic evidence. Whether this expert reading process can serve as an effective prior for ECG agents remains unclear. To address this question, we introduce LuminaECG, a clinically structured ECG reasoning framework that reformulates ECG interpretation as measurement-grounded visual reading. ECG signals are rendered on standard electrocardiographic grid paper to preserve the spatial and scale cues used in clinical reading. P-wave, QRS-complex, and T-wave boundaries are explicitly delineated, and color-coded segmentation decomposes the waveform into discrete visual measurement primitives. A general 2B vision-language backbone is then trained with low-rank supervised fine-tuning to associate these primitives with diagnostic reasoning, without architectural modification. Across open, proprietary, and ECG-specialist zero-shot baselines, LuminaECG improves both waveform measurement and diagnostic recovery. It reaches a clinically meaningful reader tier on the CODE-test benchmark, transfers across geographically diverse ECG datasets without retraining, and generates reports whose structure contains an emergent prognostic signal. These findings suggest that effective ECG agents require not only larger models, but supervision that preserves the alignment between measurable waveform evidence and clinical knowledge.

Semi-supervised adaptation of vision foundation models (VFMs) commonly freezes the pretrained backbone and updates lightweight modules such as LoRA. However, pseudo-labels have mixed reliability, and a single LoRA adapter must absorb reliable, ambiguous, and noisy gradients in the same low-rank space. This can make VFM adaptation sensitive to pseudo-label noise. We propose \textbf{TriNoL}, a \textbf{Tri}ple-expert learning framework from \textbf{No}isy \textbf{L}abels for semi-supervised VFM adaptation. TriNoL routes unlabeled samples into three confidence regions and assigns them to three LoRA experts: a Positive Expert for high-confidence pseudo-labels, an Alignment Expert for medium-confidence ambiguous samples, and a Negative Expert for low-confidence noisy samples. The VFM backbone remains frozen, and only the LoRA experts and classifier head are updated. By separating different pseudo-label reliability regions into specialized adaptation paths, TriNoL improves robustness to noisy supervision while keeping the training cost low.

Mobile app reviews are a rich, continuously renewing source of how users experience privacy and security, yet existing taxonomies of these concerns are hand-crafted and cannot keep pace with the evolving nature of the data. Automating taxonomy construction is the natural response, but scalability is the core challenge: current LLM- and clustering-based methods are developed for scientific corpora of a few thousand documents and do not extend to app review collections numbering in the hundreds of thousands. We address this gap in two ways. First, we filter app reviews for privacy- and security-related content, yielding a comprehensive corpus of over 600K reviews. Second, we introduce TaxoScale, a pipeline that handles taxonomy construction at this scale by extending an expert-defined taxonomy via Recursive Hierarchical Clustering and LLM-based node naming. TaxoScale outperforms strong automatic-taxonomy baselines on path, level, coverage, and novelty metrics, and discovers novel branches absent from prior taxonomies.

Large language models are increasingly deployed as general-purpose educational and technical assistance systems, but their underlying infrastructure does not treat languages equally. One underexamined source of disparity is tokenization: semantically equivalent content can require substantially different token counts across languages, affecting API cost, latency, and usable context length before a model is invoked. We introduce the Tokenization Equity Audit (TEA), a reproducible benchmark for measuring tokenization premiums in technical tutoring content. TEA evaluates three widely used tokenizers, GPT-4o's o200k base, Qwen2.5-7B, and Mistral-7B, on a 120-item Python debugging corpus translated from English into Bengali, Hindi, Arabic, Tamil, and Yoruba. Bengali and Hindi serve as the primary validated cases, while the remaining languages provide exploratory cross-script and cross-family comparisons. Across this corpus, Bengali requires (1.56\times) as many GPT-4o tokens as English, reducing a nominal 128k-token context window to an effective 82k-token English-equivalent capacity for the same semantic content. With the Qwen2.5 and Mistral tokenizers, Bengali requires up to (4.5\times) the English token count. Yoruba, despite using the Latin script, exhibits the highest GPT-4o tokenization premium at (2.37\times), indicating that tokenization inequity cannot be explained by script family alone. These results demonstrate that tokenization can create measurable economic and functional barriers, highlighting the need to treat tokenization as an equity-relevant infrastructure layer for underserved language communities, particularly where educational systems depend on low-cost or offline-capable AI tools.

Recent advances in sign language (SL) research have shown a trend toward unifying multiple sign language understanding (SLU) subtasks, such as isolated sign language recognition (ISLR), continuous sign language recognition (CSLR), and sign language translation (SLT), within a single framework, leading to substantial progress. Meanwhile, sign language production (SLP), which generates sign sequences from text, has also attracted growing attention. This naturally raises an important question: can sign language understanding and production be unified within a single framework? Compared with unifying SLU subtasks, this problem is substantially more challenging. Existing SLU tasks largely share the same direction of mapping, namely from sign inputs to linguistic outputs, whereas SLT and SLP lie in opposite directions of sign-text mapping. A unified framework must therefore address two key challenges: (1) bridging the modality gap between continuous sign motions and discrete text tokens through a shared sign tokenizer that supports both linguistic abstraction and motion reconstruction; and (2) learning a single conditional autoregressive model that can take either sign or text as input and generate the corresponding target sequence in the opposite modality. To this end, we propose Uni-SLTP, a unified framework for SLT and SLP with two key components: (1) a shared sign tokenizer that converts sign sequences into discrete tokens and latent representations, capturing both semantic and reconstructive information; and (2) a unified autoregressive generation model that formulates both tasks as conditional sequence generation. Experiments on widely used public datasets show that Uni-SLTP achieves superior motion accuracy for SLP while maintaining competitive SLT performance.

Continual self-evolution requires LLM agents to transform environmental interactions into reliable and reusable experience. Existing methods typically refine individual trajectories or abstract shared knowledge from related trajectories, but their experience representations are often disconnected from the underlying reasoning process. This limits feedback attribution, cross-task transfer, and update and retrieval efficiency, particularly in complex reasoning tasks with outcome-level feedback. To overcome this limitation, we propose \textbf{T}ree-\textbf{o}f-\textbf{E}xperience (ToE), a structured experience-management framework that aligns experience organization with the hierarchical reasoning process of LLM agents. Specifically, ToE organizes the experience into a shared tree of analytical perspectives and reasoning paths, whose reliability is calibrated through environmental outcomes to support systematic updating, transfer, and efficient retrieval. The experimental results on \textsc{Game of 24} and \textsc{FinEvolveBench} show that ToE substantially improves both problem-solving performance and efficiency. On \textsc{Game of 24}, ToE achieves a 31.4\% relative improvement in accuracy over the experience-free ToT baseline. On \textsc{FinEvolveBench}, ToE improves tsIC by an average of 41.24\% over the experience-free pipeline across 12 evaluation settings, whereas conventional experience-management methods often underperform experience-free baselines.

Users of modern platforms repeatedly need summaries of recent dialogue, but the window rarely contains enough context to be interpreted on its own. We formalize this setting as streaming dialogue summarization, where a system must summarize a current window using selective memory from an unbounded history under a fixed budget. We show that the central challenge is not how much history is accessed, but whether memory recovers the evidence that the current window presupposes. We construct a benchmark and evaluation protocol that separately assesses whether memory contains gap-resolving evidence and whether the generated summary reflects it. We propose ReMEMBER, a missing-evidence memory framework that conditions retrieval on unresolved window dependencies and refines retrieved chunks into evidence-dense memory under a fixed budget. Experiments on dialogues with histories up to 160K tokens show that ReMEMBER improves memory recall and gap-resolution completeness over memory construction baselines under the same budget.

Large traveling salesman problem (TSP) instances require a solver to allocate limited computation while preserving the validity of its outputs. Existing neural--operations-research (OR) hybrids predict guidance without requiring learned transitions to satisfy constraints discovered during search. DualCert introduces \emph{constraint-coupled learning}, in which current degree equations and dynamically separated subtour-elimination constraints (SECs) define each learned transition. At each refinement, the degree equations and selected, strictly satisfied SEC equations, with positive slacks, define an iterate-dependent primal-slack Karush--Kuhn--Tucker (KKT) manifold. Repaired dual variables and violated SEC rows define a local cost field. An exact constrained mirror-descent step maps each finite state to a positive state on the same manifold. Where selected rows and deterministic ties remain fixed, implicit differentiation maps parameter perturbations into the manifold tangent space and reuses the forward constraint operator for the local-cost-field derivative. The terminal edge state allocates computation across Held--Karp ascent, candidate-graph edge tests, and tour construction under a fixed budget. Deterministic verification recomputes original costs and accepts only verified candidate-graph lower bounds and edge decisions. On 1,000 held-out TSP1000 instances, DualCert attains a mean tour-cost gap of \(0.0573\%\) from Lin--Kernighan--Helsgaun version 3 (LKH-3) reference tours in \(9.55\) batch-amortized seconds per instance. It returns a verified candidate-graph lower bound for every instance and achieves \(81.46\%\) edge-decision coverage. The mean gap is \(67.1\%\) smaller than the reported NeuroLKH mean gap. Thus, optimization constraints govern learning, while deterministic verification preserves output validity.

We study decision-focused learning (DFL) in shortest-path network interdiction (SPNI) games, a Stackelberg game where an interdictor (leader) strengthens the networks' arcs against attacks, while an evader (follower) who is uncertain about costs of attacking network arcs relies on a machine-learned predictor to identify the shortest path. While DFL is highly effective as an end-to-end optimization framework, we show that it faces a fundamental structural failure when employed in this game setting: its training objective admits a broad decision-equivalence class of cost estimators that achieve zero nominal loss yet fail under interdiction, reversing DFL's usual advantage over a naive prediction-focused learning (PFL) approach. To address this, we propose Adversarial DFL (A-DFL), which replaces nominal training samples with interdicted scenarios to collapse the harmful equivalence class. Experiments on synthetic and real-world networks confirm that A-DFL restores DFL's advantage in this game setting, enabling effective end-to-end optimization.

Text-to-music generation has advanced rapidly, but current systems still rely primarily on global text prompts, leaving the structural organization of generated music implicit and difficult to inspect, control, or revise before audio generation. To address this issue, we introduce MusicLayout, an explicit intermediate representation for controlling musical structure in text-to-music generation. MusicLayout describes a musical piece as a time-aligned layout of sections, textures, repetitions, variations, and instrument-level arrangements, serving as an interpretable planning layer between textual intent and the generated music. We integrate MusicLayout into a text-to-music framework built on a unified autoregressive formulation, where the model first generates a MusicLayout representation and subsequently predicts audio tokens conditioned on this representation within a single sequence. The resulting MusicLayout can be inspected and modified prior to audio generation, providing a mechanism for layout-level structural control. We evaluate MusicLayout through layout-conditioned generation, layout manipulation experiments, and matched-data ablations, providing evidence that explicit layout planning can improve long-range structural organization and support layout-level control.

Graph neural networks typically propagate information through repeated message-passing layers, coupling the distance over which information travels with the number of nonlinear transformations applied. This coupling can make deep architectures difficult to optimize and can lead to over-smoothing, over-squashing, and the loss of long-range information. Linearized Graph Sequence Models (LGSMs) address this issue by separating information depth from processing depth and treating the successive propagation states of each node as a sequence. However, existing LGSMs construct these sequences using fixed graph operators, limiting their ability to adapt propagation to the input graph, node features, and downstream task. We introduce HOPPER, an end-to-end learnable extension of LGSM that learns how hop sequences should be extracted before they are processed by a modern state-space model. Our framework supports feature-conditioned, structure-aware, graph- and hop-adaptive propagation mechanisms while preserving permutation equivariance. Standard adjacency-based and non-backtracking LGSM sequences arise as special cases of our proposed extractor family. We show that HOPPER is state-of-the-art or competitive across the ECHO-Synth benchmark, and that varying the maximum neighborhood size of message backtracking cancellation (i.e. structural memory window) can optimize accuracy on the LRIM physics-based long-range dependency benchmark. These results demonstrate that learnable sequence extraction provides a flexible and effective approach to long-range graph representation learning.

What should a machine learning model learn when data is missing during training? We look at the learning process from a dynamical systems perspective, cast data missingness as a structured loss of actuation that limits controllability of the parameter error dynamics, and ultimately derive adaptation mechanisms with Lyapunov stability characteristics that throttle model updates in ways that preserve learning coherence under partial, intermittent observability. Under recurrent excitation, our analysis provides ISS-type residual-to-state bounds with respect to a bounded closed-loop mismatch between the loss residual and the preconditioned update geometry. We evaluate the efficacy of our directional observability-aware adaptive learning approach on multimodal contexts, reinforcing its premise in promoting learning coherence and stability even in pathologically sparse domains and problems.

Institutional policies stay in natural language while the systems that check compliance demand machine-readable constraints. Bridging that gap is still done by hand. PolicyKG closes the loop. It is an LLM pipeline that reads a policy PDF, classifies each sentence as an obligation, permission, or prohibition, lifts the label into first-order deontic logic, and emits SHACL constraints. Four stages run on a LangGraph state machine with per-stage validators. The piece that matters most is the Corpus Adapter: a YAML vocabulary registry that grounds LLM predicates in a target ontology. Retargeting to a new domain means swapping the registry, not retraining a model. On the Asian Institute of Technology Policies and Procedures corpus (1,663 sentences, 443 rules), PolicyKG reaches 86.9% deontic classification accuracy (Cohen's kappa = .709). Three annotators independently re-label a 50-item sample and agree at Fleiss' kappa = .844. SHACL shape correctness on a 69-shape subset is F1 = .866. The FOL path handles 79.2% of rules; the rest go through a direct NL-to-SHACL fallback. We audited every one of the 443 rules for second- or higher-order constructs. An automated regex checklist flagged none, and a first-author pass on the 92 FOL-fallback cases confirmed the same. The exact upper 95% Clopper-Pearson bound on the true HOL rate is 0.67%. This is an audit finding for one corpus, not a proof of FOL sufficiency for institutional policy. Swapping the AIT registry for a GDPR registry raises exact property alignment from 1/15 to 11/15 (Fisher's exact p < .001; Cohen's h = 1.53). On the LexDeMod lease-contract benchmark (N = 200), Macro F1 drops to .370 because lease English uses "shall be entitled" for permission -- exactly the vocabulary mismatch registry swap is meant to fix. Repeated runs produce hash-identical SHACL outputs.

Financial agents can turn correct context into an unauthorized effect: a customer-facing commitment, trade, or deployed policy. We present SAGE-Fin, a finance-specific authority-handoff contract that makes the proposed effect, not merely its text, the object of runtime control. SAGE-Fin compiles proposals into typed, adapter-bound candidates; records missing or stale institutional obligations as coverage debt; contracts authority under current market, account, policy, and dialogue state; and requires an exact-artifact receipt whose nominal type matches the consuming response, execution, or policy adapter. Evidence and workflow progress cannot substitute for effect authority, and prior authorization is rechecked after state changes. Across an authored 616-case catalog, five deterministic specifications yield 3,080 outputs; a label-isolated harness obtains 616/616 binary reference-prototype parity, including 3/3 named response-gate fixtures, while 22 tests cover selected paths. These results establish executable conformance, not independent safety accuracy. Separately, SAGE-Fin's response gate processed real customer-facing production requests at a confidential digital-asset platform. An operational team independent of the implementation team reached a strongly positive post-deployment conclusion on practical usefulness and workflow fit, and end-user feedback was also strongly positive. Disclosure permits only the review's independence, stakeholder classes, assessed dimensions, and directional conclusion, so this is qualitative field corroboration rather than an aggregate effect estimate. Three distinct de-identified predecessor failures, with independently confirmed 0/3 interception, ground repeated-emission drift, stale account evidence, and missing escalation state without estimating prevalence or treatment effect.

Emergency-department (ED) triage requires clinicians to rapidly identify patients who need immediate attention, determine who can safely wait, and prioritize limited clinical resources. At presentation, however, information may be limited to a chief complaint and initial vital signs. Clinically important details, including symptom onset and progression, associated symptoms, medical history, and medication use, are often obtained through focused conversation. Effective triage therefore requires clinicians to identify information gaps, ask appropriate follow-up questions, and update their assessment as new evidence becomes available. Most existing ED benchmarks evaluate acuity prediction from a fixed clinical snapshot. Although this formulation measures predictive performance after patient information has been assembled, it does not capture the interactive process through which triage-relevant evidence is elicited and interpreted. Existing medical dialogue datasets support the study of clinical communication, but dialogue statements are not always linked to temporally ordered events in the electronic health record (EHR). We introduce EHR2Dial-Triage, an agentic conversation-generation framework and benchmark grounded in MIMIC-IV-ED. The framework constructs triage conversations under explicit role-based and temporal information boundaries. Each accepted patient disclosure is linked to its supporting EHR event and the first dialogue turn at which it becomes available. EHR2Dial-Triage enables controlled evaluation of information elicitation, evidence use, five-level Emergency Severity Index prediction, and patient-facing communication across models and patient personas. It provides a structured setting for studying conversational triage as a dynamic process of clinical information acquisition, reasoning, and communication.

As generative AI increasingly becomes a common source of daily decision-making, including financial choices, it is critical to understand how people evaluate AI-generated financial advice. We conducted a preregistered vignette experiment (N = 285) in which substantive financial content---including facts, numerical values, recommendation direction, and core reasoning---was held constant while communication style varied across AI Financial Assistant (AI), Certified Financial Planner (Expert), and Online Community Forum (OC) advice. Displayed source attribution was independently manipulated through correctly labeled, unlabeled, and mislabeled conditions, allowing us to separate attribution effects from source-specific communication cues. Expert advice was rated more favorably than AI advice on 9 of 10 outcomes (|d|=0.20--0.47), and this advantage remained visible without source labels, where Expert advice outperformed AI advice on 8 of 10 outcomes (up to d=0.60). Correct labels added limited differentiation, whereas mislabeling increased ratings of AI advice for situational fit and overall quality (d=0.42 for each) and attenuated the Expert advantage in situational fit (d=-0.36). Descriptive analyses further showed that AI advice was most responsive to displayed attribution and, conversely, that advice-style differences were most visible under an AI label. These findings show that financial-advice evaluations are shaped jointly by displayed attribution and message-level communication cues. We position disclosure not as a neutral transparency mechanism, but as an interpretive frame whose accuracy and interaction with message cues can shape trust and reliance.

Node ranking is a fundamental problem in graph information retrieval, measuring the relative importance of nodes and supporting a wide range of applications such as influence analysis, recommendation, and graph-based retrieval augmented generation. However, exact computation of graph-based ranking measures is often computationally prohibitive at scale. Existing GNN-based ranking methods provide scalable approximations, but they are typically tailored to individual ranking criteria and require retraining for each downstream task, which limits their transferability and efficiency. Recent graph pre-training approaches aim to enable knowledge transfer across tasks, yet their learning objectives are largely misaligned with node ranking, resulting in suboptimal adaptability to ranking-oriented applications. To address these limitations, we propose PreGress, the first ranking-native pre-training and prompting framework for supporting a wide range of node ranking tasks. PreGress performs multi-task pre-training using our carefully designed objectives, including degree centrality prediction and attribute reconstruction, to jointly capture structural and attribute information. To support heterogeneous ranking criteria, we design lightweight, task-specific prompt modules that adapt a frozen ranking backbone to downstream tasks without full retraining. Experiments on six public graphs and two real-world query-to-item benchmarks---Yelp2018 and MovieLens-100K---together with a controlled five-criterion graph-access study demonstrate strong ranking quality with low task-specific state overhead.

Uncertainty Quantification (UQ) aims to measure the reliability of model predictions, serving as a critical safeguard for deploying Vision-Language Models (VLMs) in safety-critical scenarios. Post-hoc approaches are widely adopted due to their lightweight nature, mapping the outputs of VLMs to uncertainty measures through learnable modules or inductive summarization. However, Post-hoc approaches remain inherently confined to fitting the failure patterns of the source domain, ignoring the dynamic nature of test distributions. To address this challenge, we propose a Dynamic Distribution-Aware Uncertainty Quantification framework (DDA-UQ) that shifts the paradigm from static mapping to a dynamic distribution-aware process. During training, we leverage a Gaussian Mixture Model to model the VVLMs'embedding space and extract distributional evidence, thereby dynamically deriving uncertainty estimates. During inference, the design dynamically responds to changes in the data distribution. Extensive experiments demonstrate that our approach significantly outperforms state-of-the-art methods.

Large Language Models (LLMs) have achieved remarkable success across a wide range of tasks. However, fine-tuning LLMs for Gloss-Free Sign Language Translation (GFSLT) remains a challenge. In this paper, we investigate how to effectively adapt LLMs to the GFSLT task. We show that there are two key issues that need to be solved: (1) the inherent distributional gap between visual feature inputs and text feature inputs makes it difficult for LLMs to interpret visual inputs; and (2) existing approaches typically concatenate visual and textual features in an autoregressive framework, which leads to the model overemphasizing textual inputs and deprioritizing visual cues, as LLMs are pretrained predominantly on text-centric data. To address the first challenge, we propose a simple yet effective method named Filtered Pseudo-Gloss CTC Pretraining, which leverages filtered pseudo-gloss sequences generated from text sequences to supervise the training of the visual backbone. To tackle the second issue, we introduce a Visual-Prioritized Distillation training strategy. Specifically, we define a visual-only prediction path in which text inputs are masked, and the model is required to generate the target sequence relying solely on visual inputs. To guide this path, the outputs from the standard visual-textual prediction are then distilled into the visual-only prediction path, encouraging the model to prioritize visual features. Comprehensive experiments and qualitative analyses demonstrate the effectiveness of the proposed model. The proposed SignLlama achieves very competitive performance on multiple datasets for GFSLT tasks, without using any extra modalities or external sign language datasets for pretraining.

We prove a sharp lower bound for smooth nonconvex stochastic optimization with uniformly bounded gradient noise. In the \(K=1\) fresh-sample model, every randomized adaptive algorithm requires $$Ω\left( \frac{ΔL}{ε^2} + \frac{ΔLσ^2}{ε^4} \right)$$ queries to find a point with expected gradient norm at most \(ε\). This matches the standard upper bound and, to the best of our knowledge, resolves the question raised by [Arjevani et al. 2023] of whether almost-surely bounded oracle error permits a better rate than bounded variance. The proof was independently generated with GPT-5.6 Sol in Codex's Ultra mode during a two-hour session. The human author supplied the prompt and was responsible only forchecking the proof and revising and polishing the manuscript.

Black-box privacy scores for retrieval-augmented generation (RAG) are difficult to interpret unless the audited defense's active pipeline hook is known. We propose an active-path audit: inventory source-level hooks over retrieval, retrieved content, and generation; map each metric to the leakage channel it observes; and validate generated-text effects with exact-match canaries. In our benchmark reimplementations, the DP-style defenses modify retrieval scores only: their generation hooks are TODO-flagged stubs that return responses unchanged. This active path explains why they affect membership-inference behavior but track No-Defense on generated-text named-entity leakage, measured by NEL_strict. By contrast, the end-to-end LPRAG path is canary-validated on the email channel, recovering 53/150 canaries under No-Defense and 0/150 under LPRAG. These findings concern our reimplementations on our stack, not released defenses or defense families; the contribution is a methodology and case study, not a universal ranking

The proliferation of AI-generated images produced by Latent Diffusion Models (LDMs) has raised critical concerns regarding copyright infringement and misinformation. Although existing frequency-domain watermarking methods embed handcrafted geometric patterns into the initial latent noise prior to generation, they suffer from limited capacity and rigid pattern designs. We propose DeepFreqMark, an end-to-end learnable frequency-domain watermarking framework that replaces manual pattern engineering with a neural message encoder and decoder. To circumvent the computational bottleneck caused by Denoising Diffusion Implicit Model (DDIM) inversion during training, we introduce a Spherical Linear Interpolation (Slerp)-based attack simulation. This approach operates directly on the noise latent while strictly preserving the Gaussian variance. Extensive experiments demonstrate that DeepFreqMark achieves significantly lower Bit Error Rates (BER) than baseline methods under real-world attacks and scales to 256 bits message capacity. Our source code is available at https://github.com/chenhsiu48/DeepFreqMark.

Quantum low-density parity-check (qLDPC) codes can encode multiple logical qubits using sparse parity checks, yet searching for useful finite-length instances remains a challenging design problem because code performance must be optimized while satisfying practical constraints. Motivated by recent advances in artificial-intelligence agents for scientific discovery, we develop a multi-agent framework for discovering practical qLDPC codes. The framework combines specialist proposal and review, persistent scientific memory, long-horizon evolution of executable programs, and deterministic construction and evaluation within a closed-loop search. These programs instantiate coset-orbit balanced-product codes, providing a search space that includes bicycle and lifted-product constructions as well as non-normal subgroup actions. To incorporate practical constraints, we restrict the search to binary CSS codes with block length $n\leq400$ and overall weight $w\leq10$. Within this regime, the framework discovers codes with leading or competitive rate--distance performance in every weight class considered, with representative instances including $[[288,16,18]]$ at $w=7$, $[[288,18,18]]$ at $w=9$, and $[[234,28,18]]$ at $w=10$. The search also uncovers structurally distinct, high-performing constructions, including a $[[336,12,\leq24]]$ candidate and a $[[368,18,16]]$ code, both of which are genuine balanced-product constructions with non-normal subgroup actions. When evaluated under code-capacity depolarizing noise using a common BP-OSD decoding protocol, the discovered codes also exhibit low logical failure rates. Together, these results provide hardware-relevant finite-length candidates for further experimental evaluation and show how structured agentic search can contribute to scientific discovery.

Retrieving relevant evidence from noisy web data is challenging, particularly in sensitive domains containing incomplete reports, heterogeneous language, and irrelevant content. We present Guardian Crawler, a reproducible retrieval-first testbed for controlled experiments on knowledge discovery and evidence-grounded summarization over synthetic web-like corpora. The architecture combines BM25 retrieval with risk-aware, embedding-augmented, and hybrid reranking, followed by constrained retrieval-augmented generation with explicit document citations. Experiments on a synthetic 900-document corpus and 10 queries produced the highest descriptive retrieval scores under risk-based reranking, with P@10 = 1.00 and NDCG@10 = 0.94, compared with 0.94 and 0.81 for BM25. The best hybrid and BM25+Semantic configurations reached NDCG@10 values of 0.94 and 0.88, respectively. All 41 evaluable generated bullets passed the lexical coverage threshold; an automated LLM judge classified 36 as supported, one as partially supported, and four as unsupported. These results demonstrate the feasibility of Guardian Crawler as a controlled testbed but do not establish statistical superiority, human-validated faithfulness, or transfer to live-web investigative environments.

Public infant cry corpora are small, label-incompatible, and almost always evaluated one corpus at a time. We ask what this practice hides and what fixes it. Across four cry corpora screened by a multi-level leakage audit (byte-level and embedding-level deduplication plus a within-corpus train-test near-duplicate audit), we probe four frozen encoders and a handcrafted baseline under a unified five-class need ontology and shared task formulations. The audit exposes what single-corpus evaluation conceals: within-domain macro-F1 swings by 0.57-0.80 for the same encoder, cross-corpus transfer is negative on average (negative-transfer ratio 0.19-0.35, significant in 18 of 30 directed cells, BH-FDR), and 349 content-identical clip groups carry conflicting metadata labels across corpus distributions. The same audit, however, reveals a consistent way forward. Transfer into the noisiest corpus is consistently positive in effect size at matched training size and after near-duplicate removal, offering a practical recipe for small, noisy corpora. Frozen probes saturate at modest label budgets, while stabilized fine-tuning wins with full labels; domain-adaptive pretraining significantly beats stabilized fine-tuning at 5-10-shot (the 1-shot advantage is not robust to optimization-seed variance) but shows no significant advantage at 50-shot or beyond. In the tested binary, shared-label settings, ontology-mapped joint training wins in all four encoder-by-target combinations, whereas naively merging unmapped labels costs up to 37 F1 points. We release the ontology, mapping code, and audit pipeline, turning incompatible cry corpora into a usable joint-training resource.

Model selection for imbalanced binary classification often uses the Matthews correlation coefficient (MCC), but thresholding makes validation rankings threshold-dependent. SoftMCC is a post-training MCC validation framework on established probability-valued confusion counts, coupling an MCC-specific calibrated identity with a tie-aware, shared-pool selection protocol. Its core score is a covariance-normalized probability-label association, reduces exactly to MCC for hard predictions, and is Pearson-bounded. Under perfect population calibration it equals the Brier skill score with identical candidate ordering; outside that regime the gap does not identify calibration error. Across 18 settings with 12 duplicate-safe grouped repeats, SoftMCC attains the best stability mean rank (2.31) and highest mean tie-corrected Kendall's W (0.659), with a significant Friedman test (p=0.007); Nemenyi analysis separates it from AUPRC and MCC@0.5, while 14-source-family sensitivity retains only the latter. Selected-model utility shows no advantage. Three of six prespecified comparisons have negative mean test-MCC differences, only F1@best survives Holm correction (p=0.014), and the dataset-level test is not significant (p=0.117). Label permutation lowers mean W to 0.092; temperature scaling shifts SoftMCC rankings (mean Spearman 0.851) whereas rank-based and threshold-optimized metrics remain invariant. SoftMCC is a calibration-sensitive MCC-family selector with bounded stability and utility evidence.

Interactive video world models generate rollouts autoregressively under an action stream, yet they are trained and evaluated almost exclusively on factual prediction. We study counterfactual generation inside the rollout: given a trajectory the model has itself generated, what would have happened had the actions differed from step t* onward? We formalize noise-coupled twin rollouts --- a factual and a counterfactual branch sharing the generated prefix and the future exogenous noise sequence, diverging only in the action stream at an intervention point. Because the factual branch is self-generated, its exogenous noise is known exactly: the abduction step of Pearl's counterfactual procedure is exact by construction, sidestepping the approximate-inversion problem faced by editing-based pipelines. Noise coupling further turns the minimal-change principle into a per-sample verifiable property: we define a spatiotemporal locality metric that penalizes divergence outside the causal descendants of the intervention, computable against simulator ground truth without a learned judge. Forking the simulator state at t* yields ground-truth counterfactual re-renders, which we use as verifiable rewards for post-training. This note establishes the formal framework, metric definitions, and positioning; experiments are forthcoming.

This work focuses on the impact and detection of clear contact lenses in the context of iris recognition. While the detection of cosmetic or patterned contact lenses has been extensively studied under the presentation attack detection (PAD) paradigm, clear prescription contact lenses, that are typically transparent, have received comparatively less attention despite their widespread use. Unlike patterned lenses, clear lenses introduce no salient texture artifact, making them difficult to detect and are often assumed to have no impact on iris recognition. We first examine this assumption using the commercial VeriEye matcher on four benchmark datasets and show that clear lenses marginally degrade genuine match scores and increase verification error. We then propose a two-stage contact-lens detection framework. Stage~1 uses an existing PAD model to identify patterned lenses, while Stage~2 focuses on the more challenging clear-lens versus no-lens distinction using a ConvNeXt-Base model equipped with Mask-Guided Spatial Attention (MGSA). The proposed MGSA module incorporates a Hough-derived anatomical ROI mask together with learned spatial attention and Squeeze-and-Excitation channel recalibration, allowing the network to focus on subtle limbal cues associated with clear lens wear. Across four datasets, the full pipeline consisting of both patterned and clear contact lens detection achieves between 90.0\%--98.8\% accuracy. Finally, we introduce a z-score calibration method that adjusts VeriEye match scores when a clear lens is detected in the input images. This calibration reduces EER by 4.1\%--28.3\% across datasets, demonstrating that reliable clear contact lens detection can directly improve iris verification performance.

Parkinson's disease (PD) is the second most common neurodegenerative disorder. Typical machine learning screening methods require PD labels, but the available data is limited by privacy concerns and the need for expert annotation. We propose a label-free face-plus-voice PD screen built entirely on frozen pretrained encoders--a face-expression Vision Transformer and HuBERT--in which no PD label touches any fit; the reference is training controls only. The voice modality uses a synthetic-dysarthria contrastive activation addition (CAA) direction built from time-stretch and breathy degradation of healthy speech; the face modality uses a k-nearest-neighbor anomaly score to the control embedding cluster. We introduce the alignment principle, a post-hoc analysis showing that a synthetic-degradation CAA detector works when the cosine similarity between the synthetic and real disease directions exceeds zero. Measured on the YouTubePD benchmark, this cosine is +0.37 for voice (CAA works, AUROC 0.765) and -0.48 for face (CAA fails; anomaly succeeds, AUROC 0.751). Equal-weight late fusion reaches AUROC 0.802 (95% CI [0.70,0.89]) with NPV 0.95, supporting a rule-out triage interpretation. An overfitting audit shows the voice detector transfers cleanly, while the face-side--and thus fused--AUROC is potentially optimistic pending external validation.

As large language models increasingly participate in scientific evaluation, we investigate a potential form of reward hacking: how rhetorical choices shape AI-review judgments when reported scientific content is preserved and how these effects vary across evaluation conditions. We construct a controlled corpus of 4,200 full-paper manuscripts derived from 120 anonymized ICLR 2026 submissions. Two LLM rewriters transform six rhetorical dimensions in opposing directions, and five LLM reviewers evaluate the resulting manuscripts under standard and strict protocols. We also test joint, recursive, and reviewer-guided rewriting. Our results show that rhetorical sensitivity is structured rather than uniform. Evidence framing and novelty stance produce the largest positive-negative contrasts in overall assessment, with scope framing forming a weaker second tier; the remaining dimensions have smaller or less stable effects. This hierarchy persists across human-assessed quality levels, but score movement depends strongly on the AI reviewer's original score: lower scores tend to rise, higher scores tend to fall, and directional contrasts are clearest in the middle ranges. More elaborate workflows do not reliably yield larger gains. Joint rewriting is strongly rewriter-dependent, reviewer guidance does not consistently outperform an unguided second pass, and repeated rewriting yields diminishing, configuration-dependent returns. Across conditions, the rewriter primarily determines the separation between opposing variants, whereas the reviewer determines the magnitude and sign of their score effects. Strict review lowers mean OA by 1.36 points without consistently changing rhetorical sensitivity. These findings identify when rhetorical presentation influences AI scientific review and motivate evaluation systems robust to content-preserving variation in scientific writing.

Microservice root cause analysis (RCA) requires correlating failures across heterogeneous telemetry within complex service dependency graphs. Existing methods often rely on a single telemetry modality; recent LLM-based approaches can suffer from unconstrained exploration and hallucination; and most systems stop at fault ranking without producing actionable incident response. We present GALA+, a graph-augmented LLM agentic framework centered on graph-guided investigation, which uses service dependencies to bound exploration and refine diagnosis through localized multi-modal evidence. For initial hypothesis generation, GALA+ combines complementary telemetry signals with STRIX, a novel trace- and graph-structure-aware scoring module. GALA+ then produces ranked diagnoses, incident summaries, and stratified action recommendations. We further introduce SURE-Score, a human-guided evaluation framework co-developed with industry SRE experts for assessing RCA-specific output quality beyond conventional text similarity metrics. On two microservice benchmarks, GALA+ consistently achieves the strongest overall results, surpassing the best LLM-based baseline by more than 25 percentage points in AC@1, while also receiving the highest ratings from both SURE-Score and independent human SRE evaluation.

Underwater images often suffer from diverse and coexisting degradations, including color distortion, scattering haze, texture attenuation, and uneven illumination. These degradations vary across regions and may coexist locally, making conventional uniform restoration difficult to adapt to different degradation patterns. To address this problem, we propose Coexisting and Region-wise Degradation for Underwater Image Enhancement (\textbf{CoRe-UIE}), a degradation-oriented expert collaboration framework. CoRe-UIE combines a content-preserving shared expert with four shared-backbone routed experts for color correction, scattering suppression, texture recovery, and illumination protection. The routed experts share the same architecture but have independent parameters, and are assigned to different regions through input-derived degradation cues and region-adaptive Top-\(k\) routing. We further introduce a Hilbert--Schmidt Independence Criterion (HSIC)-based representation constraint to reduce statistical dependence among expert features and alleviate redundant expert responses. Experiments on UIEB, LSUI, and U45 demonstrate that CoRe-UIE achieves competitive quantitative performance and visually balanced enhancement under diverse underwater degradation conditions.

The generation of mathematically precise diagrams from tex- tual prompts has emerged as a critical yet underexplored capability of Large Language Models (LLMs). This has been of interest to researchers in the areas of curriculum preparation, automated ranking of problem sets, and scientific publishing. For LLMs to achieve this, it requires per- fect coordination between Spatial Reasoning, Mathematical Reasoning, and Rendering systems. While existing benchmarks such as MathVision, MathVista are built for Math Reasoning or DiagramGenBenchmark, Mer- maidSeqBench on general purpose diagram generation, no prior work provides a standardized set of prompt, image pairs that can be used to evaluate the LLMs specifically on math diagram generation. This includes fields that span both both text-to-code and text-to-image paradigms. We introduce Math-Vision Diagrams, the first benchmark specifically designed to evaluate LLMs on mathematical diagram generation, and the first to assess text-to-code and text-to-image generation paradigms together in a single unified setting, agnostic of the underlying coding lan- guage or model type. Building on the Math-Vision benchmark, we select a subset of 2920 images out of 3040 from high-quality competition problems with essential visual context. A novel pipeline combining an ensemble of LLMs with Subject Matter Expert (SME) curation is presented, together with a suite of evaluation metrics. Testing several leading models against this benchmark, we demonstrate that LLMs struggle with math diagram generation. All code, data, curation pipeline, and evaluation scripts will be fully open-sourced.

Generalized Category Discovery aims to recognize known categories while identifying novel ones within unlabeled data. Existing methods, typically based on self-supervision and contrastive learning, often struggle to capture fine-grained distinctions, relying on superficial visual cues rather than the intrinsic attributes humans use for categorization. We introduce Fourier Self-Supervision, that leverages the Fourier transform of images to enhance the discrimination of subtle differences and support the discovery of new categories. Our method employs a dual frequency filtering strategy: a low-pass filter first extracts broad, abstract attributes that capture high-level category information, while a high-pass filter emphasizes fine details such as edges and textures that are essential for fine-grained recognition. Each operates on a dedicated latent space, and their overlapping representations together yield a richer, more complete feature space. This dual-frequency approach not only refines feature extraction to identify novel categories, but also strengthens the model's discriminative power in fine-grained category discovery. Experiments on multiple fine-grained datasets show that incorporating Fourier Self-Supervision outperforms state-of-the-art methods, even when the number of classes is unknown, demonstrating its effectiveness for Generalized Category Discovery. Our code is available at: https://github.com/SarahRastegar/FourEx.

AI training's rising resource intensity is straining electricity supplies and carbon budgets, motivating systematic study of memory-efficient training on constrained hardware. We benchmark five gradient optimizers (SGD, Adam, Adagrad, Adadelta, and Conjugate Gradient Descent) under three memory strategies (standard training, gradient checkpointing, and gradient accumulation) across four transformer architectures (ViT, ModernBERT, Llama 3.1 1B, and NanoVLM), measuring training loss, GPU utilization, training time, and memory usage. Gradient accumulation emerges as the most reliable strategy, cutting training loss by roughly an order of magnitude on the vision-language model and about four-fold on the language model without additional GPU memory. Contrary to common practice, Adam is not universally superior: Adadelta and SGD outperform it on the encoder and autoregressive architectures. Gradient checkpointing's effect is strongly architecture-dependent, improving vision transformer loss while severely degrading the encoder model, and it increases training time by up to 60% on memory-bound models. GPU utilization is governed primarily by architecture, ranging from 8-15% for the memory-bound language model to 96-99% for compute-bound vision models. These findings provide practical guidelines for optimizer and gradient-strategy selection in resource-efficient model training and deployment.

Multi-step language-model agents repeatedly process growing interaction histories, leading to substantial context costs. Vision--text compression reduces these costs by rendering history as images, but the resulting modality shift creates a marked capability gap. Through controlled evaluations of history recovery, matched-state decisions, and complete trajectories, we show that this gap cannot be explained by OCR quality alone. Visual-history agents exhibit systematic drift in action selection, query formulation, stopping, and evidence use, revealing an agentic policy gap. We introduce \textbf{CAPS}, a two-stage \textbf{C}ross-modal \textbf{A}gentic \textbf{P}olicy \textbf{S}elf-distillation framework that uses the same model's stronger text-history policy to supervise its visual-history counterpart. Offline trajectory self-distillation transfers successful text-policy behavior to visual-history inputs, while online policy self-distillation provides dense supervision on states visited by the visual-history policy during reinforcement learning. On SearchQA, CAPS improves over AgentOCR by 5.0\% and 3.4\% with 3B and 7B backbones, respectively. On full-history ALFWorld, the corresponding gains are 15.6\% and 14.5\%. Across settings, CAPS reduces average memory-context cost by up to 63.3\% and peak cost by up to 83.4\% relative to matched text-history policies. These results show that explicit cross-modal policy self-distillation can preserve agent capability under vision--text compression. Our code will be made publicly available in a future release.

Gravimetry images subsurface density contrasts associated with geological structures, geothermal systems, and intrusive bodies. Recovering a three-dimensional density model from gravity observations is highly ill-posed because of its non-uniqueness, limited data coverage, and the attenuation of the gravity field with depth. Classical inversion methods rely on explicit regularization and parameter tuning, whereas supervised deep-learning approaches require representative gravity--density pairs that are rarely available. This paper proposes an unsupervised depth-aware implicit neural representation for 3D gravity inversion. The density volume is represented by multiple coordinate-based neural networks assigned to overlapping depth slabs and optimized directly from the observed gravity measurements through the sensitivity matrix. Slab-specific Fourier features, physics-based depth gains, and scheduled regularization provide structural priors without requiring labeled density models. Experiments on four synthetic scenarios show that the proposed method provides better overall performance in terms of RMSE, PSNR, and SSIM than the evaluated conventional and neural baselines. It also recovers more compact and spatially coherent density bodies, improves the separation of nearby anomalies, preserves internal structures, and reconstructs their vertical extent better. These results indicate that the proposed depth-aware formulation helps to mitigate the depth ambiguity inherent in gravity inversion. In the field experiment, where no ground-truth density model was available, the method produced compact, separated, and vertically coherent anomalies consistent with the observed gravity pattern.

Tree Search-based test-time scaling of LLMs is a powerful tool for automated scientific coding. However, pure Tree Search sometimes struggles with systematic exploration, becoming trapped in local optima, or unproductive loops, especially in the vast search space of scientific methods. To address this limitation, we propose Idea Search, a framework that systematically integrates a dynamic "Idea Bank" into Tree Search. Idea Search involves three steps: (1) decomposing existing methods into atomic ideas, (2) sampling from this bank of ideas to guide branches of code mutations, and (3) dynamically updating the bank with new ideas discovered through execution. On single-cell RNA-sequencing (scRNA-seq) batch integration, Idea Search reliably breaks the plateau of a strong pure Tree Search baseline, improving the mean score from 0.678 to 0.697 and reaching a best score of 0.728. We then characterize which design choices drive these gains: bank augmentation helps bandit sampling but not random sampling, "Exploratory" prompting that prioritizes new ideas surfaces the rare best-performing solutions, while increasing sampling-level exploration is counterproductive.

Ensemble forecasting aims to sample the conditional distribution of outcomes; whether AI forecast ensembles do this correctly in a joint sense remains largely untested. We train a diffusion model for probabilistic subseasonal coastal sea level forecasts at eight US East Coast tide gauge stations, with sea level derived from reanalysis, and find that marginal and joint forecast quality decouple: positive skill at every station and lead time marginally, while joint spatial structure is worse than climatological draws. A shuffle-based permutation decomposition reveals this failure is invisible to the energy score but detected by the variogram score. Lorenz-96 experiments across 0.7-170 equivalent years show the gap persists regardless of training volume and is reproduced by a linear baseline, indicating structural inadequacy of the learned distribution. A dynamical ensemble does not replicate the failure while a deterministic emulator does, suggesting it is specific to learned emulators rather than ensemble forecasting generally.

Autonomous coding agents powered by large language models can now generate code patches directly from bug reports, but a fundamental gap remains: once a patch is produced, no mechanism independently verifies whether it truly resolves the reported problem. Prior work has sought to address this through iterative self-refinement and inference-time scaling, but these approaches either review the patch under the same interpretation that produced it or broaden candidate generation without verifying individual patches, and neither provides an explicit verification signal for assessing patch correctness. We propose RETRACE, a training-free post-generation verification framework that derives such a signal through bidirectional reconstruction and reconciliation. When a coding agent generates a candidate patch for an issue, RETRACE performs forward reconstruction to build an explicit repair rationale from the issue and the agent's trajectory; backward reconstruction then independently infers, from the patch and its trajectory alone and without access to the original issue, a description of the problem the patch appears to address, and compares this reconstruction against the original issue to produce an alignment verdict; a reconciliation stage then checks the consistency between the forward rationale and the patch, diagnoses the source of any misalignment, and either submits the patch or produces targeted revision guidance. Evaluated on SWE-bench Verified with two backbones (GPT-5-mini and MiniMax-2.5), RETRACE lifts Pass@1 by 7.0% and 3.6% respectively on the mini-SWE-agent scaffold, and delivers comparable gains on OpenHands without modification. Ablation experiments show that both the forward and backward stages contribute to the overall improvement and that adding reconciliation yields further gains.

Current hazard detection systems in autonomous driving may develop mesa objectives, learned internal goals that achieve high training performance through spurious correlations rather than genuine hazard recognition. We investigate whether human gaze patterns, captured via webcam-based eye tracking (WebGazer.js), can serve as privileged information to constrain mesa-objective formation. We collected 137,663 frame-level gaze samples synchronized with hazard annotations across 388 real dashcam clips, then test this hypothesis across two calibration protocols (9-point/45-click and 11-point/440-click), two model architectures (Random Forest and causal Transformer), and five random seeds per experiment with paired t-tests. No experiment yields a statistically significant improvement from gaze (p = 0.919, 0.578, and 0.667 respectively). A geometric analysis reveals the root cause: WebGazer's reported error (~130-257 px depending on configuration) exceeds 93% of detected hazard object sizes (median 36 px), rendering object-level gaze attribution physically impossible at this instrument precision.

A failed retrieval-augmented generation (RAG) answer can be consistent with several unseen responses to evidence repair. We introduce Pair-ID, an offline audit that holds one query, retrieval state, and reader constant, then crosses two operations, adding missing support and deleting verified nonsupport, to measure a same-failure counterfactual response vector. A complete funnel over 19,981 benchmark queries identifies 11,105 eligible Qwen failures, from which a prospectively fixed SHA-256 ordering selects 1,200 before generating any sampled response. Among 1,190 regenerated-valid failures, support addition repairs 197/600 JOINT cases (0.328, 95% CI [0.292, 0.367]), and deletion repairs 162/1,190 cases (0.136, 95% CI [0.117, 0.155]); length- and position-matched shams retain semantic contrasts of 0.223 and 0.101. The original view carries partial predictive signal for individual response cells (macro AUROC 0.678; Brier 0.152 versus 0.160 for a marginal baseline), but exact-vector accuracy, 0.637, does not exceed the 0.646 majority-vector baseline, and vector macro-F1 is 0.170. Across four readers, both marginal sensitivities recur, while pooled exact-vector agreement is 0.675-0.765 and JOINT-only agreement falls to 0.538-0.691. These results show that evidence sensitivity occurs at meaningful rates in the hash-selected eligible-failure sample, is only partially predictable from the observed failure, and is conditional on the reader. The evidence supports a frame-scoped offline response audit, not an information-theoretic impossibility result, reader-independent taxonomy, or runtime repair policy.

Large language models (LLMs) are increasingly integrated into healthcare, education, public services, and everyday decision making. They should provide comparable assistance regardless of a user's literacy, communication style, or prompt-engineering expertise. However, existing research on prompt robustness primarily focuses on adversarial attacks, prompt injection, and prompt optimization, while overlooking whether semantically equivalent requests receive different responses simply because they are phrased differently. We refer to this accessibility challenge as "Prompt Privilege": users with greater prompting expertise systematically obtain better model performance despite expressing the same underlying intent. To address this problem, we present a unified framework for measuring and mitigating accessibility disparities in LLM interactions. We introduce Prompt Equity Score (PES), a quantitative metric for evaluating performance consistency across user populations, and Prompt Equity Transformer (PET), an LLM-based agent that automatically transforms user requests into semantically equivalent, accessibility-oriented prompts while preserving their intent. PET shifts prompt optimization from the user to the AI system, functioning as an intelligent accessibility layer between users and foundation models. Experiments on the MedQA benchmark demonstrate measurable prompt privilege, with statistically significant performance disparities between low-literacy and expert-prompting cohorts. Applying PET eliminates these disparities while preserving semantic fidelity, demonstrating that accessibility-oriented prompt normalization can improve equitable AI access. By introducing prompt privilege as a new dimension of AI accessibility and PET as a practical solution, this work advances system-centered accessibility and provides a foundation for more fair, trustworthy, and inclusive AI systems.

Operational Design Domain (ODD) specifications describe where an automated driving system (ADS) is permitted to operate, but they do not prescribe what the ADS must demonstrably do once deployed within that domain. This gap between operating condition specification and behavioral validation represents a critical unresolved challenge in ADS safety assurance. This paper presents a structured, standards-grounded taxonomy of 21 behavioral competencies organized across three operational domains-Highway (HWY), Urban (URB), and Hub (HUB)-derived systematically from the PEGASUS six-layer model-based ODD. Each behavior is decomposed along longitudinal and lateral control axes and characterized against a four-property framework: Safety (gap maintenance, conflict avoidance, kinematic stability), Compliance (legal rules and behavioral norms), Comfort (rider dynamics and trust), and Efficiency (mission completion and product-level metrics). We further demonstrate that the crossing of ODD layer parameterizations with behavioral competency specifications yields concrete scenario families suitable for systematic behavioral testing and SOTIF coverage evidence. The taxonomy is grounded in AVSC00008202111, SAE J3237, and SAE J3016, and is validated as an operational specification layer through its deployment in a rule-enforced trajectory optimization system. The Hub domain is identified as a structurally distinct, underspecified domain warranting dedicated research attention.

The rise of autonomous AI agents represents a major paradigm shift in how users interact with mobile devices. Frameworks such as MobileRun and Mobile-Use can autonomously navigate Android applications and execute complex multi-step tasks. To interpret user interfaces, these frameworks rely primarily on Android accessibility (A11y) trees and secondarily on visual screenshots. In this paper, we demonstrate that this architectural dependence on unsanitized accessibility metadata, together with visual input, introduces a systemic vulnerability to indirect prompt injection. We show that adversarial prompts can cause autonomous agents to abandon their original objectives, violate context boundaries, and perform unauthorized device actions. Our empirical evaluation demonstrates goal hijacking, context drift, and unauthorized actions across visually hidden and fully exposed attack scenarios. In aggregate, MobileRun reaches an attack success rate of 0.822 with Gemma4:31B, while Mobile-Use with Qwen3.6:35B reduces this to 0.150 but does not eliminate context drift or unauthorized actions. These findings reveal that current mobile agent frameworks fail to enforce semantic context boundaries, treating passive environmental text as trusted instructions. Finally, we present a taxonomy of these attacks and discuss the need for zero-trust input validation, dedicated security agents, and strict context isolation within mobile agent architectures.

This work presents a unified multimodal AI system for damage assessment that integrates retrieval-augmented generation (RAG) models, thermal spectrum perception, vision foundation model pipelines, and exploratory wireless signal sensing. A RAG component is developed to ground a locally hosted language model in project-specific documentation, including specialized damage level classification criteria to mitigate hallucinations during inference. Controlled comparisons against static few-shot prompting demonstrate that dynamic retrieval improves grounding and factual consistency. We further compare vector-based RAG with a knowledge graph variant constructed via entity-relation extraction, and show that graph-based retrieval produces stronger responses for damage assessment queries requiring cross-document reasoning, motivating hybrid dense, sparse, and graph-aware retrieval. To address limitations of EO imagery under adverse lighting and weather conditions, infrared (IR)/thermal sensing is employed for object detection and segmentation. Our detectors generate candidate detections, yielding improved segmentation of a broad array of objects. Paired IR versus visible spectrum tracking experiments reveal failure modes, motivating multimodal fusion for robust object detection and damage analysis. Vision foundation and vision-language models are leveraged to generate synthetic damage imagery and classify damage severity with high accuracy, supporting training and validation of downstream damage assessment models. Finally, exploratory Wireless-based sensing demonstrates potential to detect presence, motion, and post-event environmental changes where EO and IR sensing are ineffective.

Biomedical research has accumulated vast molecular, clinical, and population data, yet translating this wealth into actionable knowledge remains constrained by technical and organizational difficulties. This article presents a unified treatment of two perspectives on biomedical knowledge infrastructure. The first introduces the biomedical domain to software engineers: it explains why knowledge graphs (KGs) are the central integrative data structure in modern biomedicine, characterizes five data harmonization challenges (identifier mapping, entity resolution, schema alignment, evidence integration, and provenance tracking), surveys application domains from drug discovery to digital twins, and profiles six representative KG systems with contrasting choices. The second perspective asks why engineering biomedical knowledge infrastructure remains so difficult. We argue that a contributing root cause is limited adoption of software tooling and practices that make development in other mature domains - particularly web engineering - reliably composable and reproducible: package management, typed namespaces, canonical interchange formats, service composition protocols, reproducible pipelines, and lifecycle governance. Against this backdrop, eight open engineering challenges for biomedical data integration are catalogued, each with partial solutions but no universally adopted stack. Crucially, the article shifts emphasis from describing deployed KG instances toward the reproducible process of assembling them: reusable build pipelines, versioned dependencies, and engineering practices that let others compile and customize a KG from source rather than consuming a static artifact. Together, the two perspectives provide domain grounding for newcomers and a research agenda for software engineers seeking to make transformative contributions to biomedical knowledge infrastructure.

Neuroimaging and genetic testing are two important clinical references for nervous system diseases, offering complementary diagnostic information. However, integrating genomic and neuroimaging data for precise disease diagnosis is challenging due to cross-modality heterogeneity. Existing imaging-genetics approaches mainly encode genetic information as hard-coded labels, which lose the local sequence context around disease-associated variants. To address this limitation, we propose GeneFuse, a multimodal learning framework that aligns genetic representations from pre-trained Genomic Language Models (GLMs) with features extracted from images. GeneFuse integrates two components: (1) Genotype-Conditioned Feature Modulation (GCFM), a FiLM-inspired module that uses genomic embeddings to modulate image feature maps; and (2) Uncertainty-aware Genomic Residual Fusion (U-GRF), a fusion strategy that uses imaging-derived predictive uncertainty to gate the contribution of genotypic features. We evaluate GeneFuse on early cognitive decline identification (NC vs. MCI) and dementia screening (NC vs. AD). In the APOE-centered setting, GeneFuse achieves AUROCs of 0.77 and 0.83, outperforming existing imaging-genetics fusion methods. These results indicate that GLM-derived genomic embeddings provide additional information to imaging.

Transformer layers generate state-dependent interaction networks: token representations determine the attention matrix, which in turn updates the representations. We study this feedback in a minimal normalized self-attention dynamics and identify the overlap gap as the central quantity governing its attractor structure in the thermodynamic limit. When tokens form internally aligned clusters and their similarity to members of the same cluster exceeds that to every other cluster by a nonvanishing amount, inter-cluster attention is exponentially suppressed as the dimension increases. This mechanism produces a high-dimensional manifold of clustered fixed points, ranging from a few macroscopic clusters to extensive microscopic fragmentation, and also controls their stability against perturbations. Starting from an unstructured Gaussian state, we find that clustered states nucleate from the diffuse background only above a finite threshold in attention sharpness, giving rise to a dynamical attention-condensation transition.

Perioperative risk prediction models are often limited by narrow surgical populations, incomplete intraoperative data, poor calibration, and limited interpretability. We present a domain-structured ensemble framework for perioperative outcome prediction using routinely collected electronic health record (EHR) data. Predictors are organized into patient-related, surgery-related, and anesthetics-related domains. Domain-specific gradient boosting models generate independent risk estimates that are integrated through a logistic regression meta-learner. We demonstrate the framework using postoperative delirium (POD) in a case-control sample of 5,386 surgical encounters (2,693 cases, 2,693 controls) from a statewide health information exchange. POD required both delirium-related ICD codes and a positive Confusion Assessment Method screening within seven postoperative days; patients with preexisting dementia were excluded. The stacked meta-learner achieved AUROC 0.899 (95% CI: 0.891-0.906), precision-recall AUC 0.881, and Brier score 0.126, compared with AUROC 0.849 for the best single-stage model. Domain ablation showed improved discrimination and calibration over a surgery-only model (AUROC 0.879, Brier 0.140). Temporal validation on held-out post-2017 data yielded AUROC 0.915. Calibration was excellent, with intercept -0.006 (95% CI: -0.083 to 0.070) and slope 1.035 (95% CI: 0.982 to 1.088). Decision curve analysis, corrected for case-control sampling, showed positive net benefit across clinically plausible thresholds. The modular framework supports alternative outcomes, extension of predictor domains, and dynamic risk updating, providing a scalable foundation for interpretable, calibration-aware perioperative clinical decision support.

During standard radiotherapy planning, repeated CT acquisitions are often required for patient registration, verification, and adaptive planning, resulting in increased cumulative X-ray dose. To mitigate this, low-dose cone-beam CT (CBCT) is routinely acquired during treatment delivery. However, CBCT image quality remains insufficient for accurate dose calculation and adaptive radiotherapy planning due to increased scatter, noise, beam hardening, and reconstruction related artifacts. This study develops a supervised deep learning based CBCT to CT synthesis framework using a conditional denoising diffusion probabilistic model (DDPM), where the generation of a CT-based planning for accurate positioning and dose calculation is obtained using generative models with low dose CBCT imaging. Beyond demonstrating CBCT to CT synthesis, the primary objective is to investigate how the representation of CBCT input data, either standard clinical DICOM CBCT images or filtered back-projection (FDK) reconstructions from raw projection data, affects the performance of diffusion based CT synthesis. The overarching aim is to assess whether physics aware CBCT representations better support CT-equivalent image quality while maintaining reduced imaging dose in radiotherapy workflows.

Situated language use is multimodal and embodied. For example, gestures can carry information that is absent or underspecified in the speech signal, yet dialogue models typically rely on transcripts alone. We study how much referential information gestures and their combination with speech carry in multimodal dialogue under different partner visibility conditions. % We build models that identify the intended referent in a video-mediated referential communication game based on either the speech transcript, the skeletal representation of gesture, or both modalities. Our results show that gesture alone is predictive of the intended referent and that multimodal fusion is most beneficial when the transcript-based model is uncertain. Training-only alignment of learned representations with the referent image further improves the fusion model performance. % In a comparison with human interaction data, we further see pragmatic effects of interlocutor visibility on gesture production and informativeness as well as an entrainment effect in speech and multimodal, but not gesture, performance across rounds of repeated interaction. We thus make contributions to the technical modelling of multimodal information in human dialogue and the analysis of human interaction data via trained model representations.

Recent advances in Passive Acoustic Monitoring (PAM) offer an opportunity to obtain ecological spatial point-process data at unprecedented scale. However, realizing this opportunity necessitates the development of accurate and scalable localization methods. In real-world outdoor soundscapes, however, the assumptions underlying classical localization methods such as hyperbolic and score-based localization are routinely violated by multipath dominance, near-field effects, and complex propagation. Under these conditions, classical localization methods become brittle, with extreme errors possible even in small detection arrays. Rather than statistically replacing the underlying physics, we propose a method to refine it and increase robustness outside of ideal operating conditions: a learned model operating on physics-informed acoustic features corrects a fast hyperbolic solver where it produces implausible solutions, substantially reducing catastrophic worst-case errors while matching its median accuracy on field data. We further provide calibrated, geometry-aware uncertainty estimates suitable for propagation into downstream spatial models. Evaluating on distributed microphone arrays in real and simulated outdoor environments, we demonstrate that the proposed method yields robust, uncertainty-aware localization, providing a step toward scalable automated wildlife monitoring in complex acoustic environments.

PTQTP decomposes LLM weight matrices into two ternary (trit) planes with two free per-group scales. Tying the scales to a fixed ratio of three collapses the decomposition into a single uniform nine-level quantizer, a known balanced-ternary identity. To our knowledge, at the time of writing, this work is the first to impose that identity as a constraint inside PTQTP's solver. The two trit planes then fold losslessly into one 4-bit code plane that we make the persistent serving representation: disk bytes, expert-cache bytes, and kernel input are the same 4.0625-bits/weight blocks, consumed in one integer dot pass. For this conjunction (ratio-3 nine-level code, CPU-SIMD kernels, SSD expert streaming, identical persistent bytes) we likewise found no precedent. We apply this to the routed experts of DeepSeek-V4-Flash-0731, a 284B-A13B mixture-of-experts model, quantizing in one shot from the released MXFP4 expert weights and streaming experts from SSD on a 64 GB laptop. Against a 4.5-bit Q4_K baseline, measured one process per fixture with an expert-lossless anchor arm as reference control, the tied model matches the official serving API on 5/5 fixtures at step 0 (Q4_K: 4/5) and 12/14 captured continuation steps (11/14), scores 86 vs. 84 on a 100-item MMLU subset, decodes 6.7% faster in decode phase, and ships 9% smaller files: no detected fidelity difference at these small evaluation sizes, and every fixture-level difference between the arms traces to a single measured near-tie cell. The tied fit nevertheless shows higher weight-reconstruction error and worse perplexity, a measured dissociation between proxy metrics and reference fidelity. A cumulative trunk-ternarization ladder and bitwise-pinned aarch64/x86-64 kernels complete the report. All code, formats, and evaluation artifacts are open source in the fucina inference stack.

Thinking with images allows a multimodal model to compensate for limited perception by invoking visual tools through code. Yet the prevailing SFT-then-RL recipe creates a different supervision misalignment at each stage. SFT is expected to teach how to use tools, but trajectories from stronger teachers may succeed through perceptual capabilities that a smaller student cannot reliably reproduce or exploit, causing the student to imitate tool-call patterns without learning how to make them useful. RL is expected to teach when to use tools, but outcome-only rewards make fallible tool execution a liability and suppress tool use, whereas a blanket bonus for every correct tool-using trajectory encourages valid but ineffective operations. To address these two misalignments, we introduce ToolVision. During SFT, a multi-agent pipeline explores candidate trajectories, and a committee including student-scale models scores stepwise evidence gain to rank and prune the search branches. Only successfully executed trajectories with correct final answers are retained for SFT. Before RL, ToolVision compares the learner's performance with and without tools, then rewards successful tool use only on questions where tools provide a clear benefit. Both signals are constructed automatically from public task data without additional human annotations of tool use or necessity. ToolVision-8B improves over its base on all seven main benchmarks, surpasses Thyme-7B, CodeVision-8B, and CodeDance-7B on all three high-resolution benchmarks, and outperforms Qwen3-VL-32B-Thinking on V* and HRBench 8K. We will publicly release the datasets and source code.

Outlier detection in decentralized data environments is a challenging task for many machine learning implementations, particularly in settings where data cannot be shared. Recently, there have been advances in federated outlier detection, some of which are based on the use of autoencoder networks. The introduction of attention mechanisms to autoencoders boosts their efficiency. However, the application of attention-based models in federated learning remains underdeveloped due to the absence of proper aggregation functions for these types of networks. In our work, we propose two novel aggregation functions tailored for attention-based autoencoders, which better preserve the learned information stored within the memory modules of these networks. We evaluated our approach on the KDDCUP10 dataset, and we showed that the proposed methods achieve up to 2.9\% and 5.1\% better results for F1 score and AUC ROC respectively when compared to traditional autoencoders.

How much of a vision-language model's (VLM) spatial understanding remains after the action post-training process of building a vision-language-action model (VLA)? We probe depth perception, a primitive of spatiogeometric understanding, from every decoder layer of a weight-matched open-source base VLM/VLA pair: Molmo2-ER and MolmoAct2-LIBERO. First, the VLA decodes depth worse at every layer, a persistent gap we call the floor. Second, the degradation is not uniform: while the base VLM's depth decodability improves through its final layers, the VLA's collapses, an additional late-layer drop we call the cliff. We causally localize the cliff to late-layer MLP interference: ablating the late-layer MLP writes recovers the majority of the terminal decodability cliff, while matched attention ablations and the same intervention in the weight-matched base VLM produce no comparable recovery. A module-level decomposition explains this dissociation: the base VLM carries depth most accessibly in accumulated MLP writes, whereas action post-training collapses depth decodability in the late accumulated writes.

Characterising optimisation problem instances is a fundamental part of understanding the behaviour and performance of different algorithms as well as providing information for algorithm selection and configuration. In this paper we propose a novel approach to problem characterisation based on the representation of instances when implemented as a program. The intuition is that the complexity of the code required to express an objective function should relate to the complexity of the search landscape. We identify the Halstead volume as a measure of code complexity, which can be seen as a simplified version of the entropy of the program. Given a code implementation of the objective function, the Halstead volume and entropy can be quickly calculated using existing libraries. We apply the proposed complexity measures to the well-known BBOB optimisation problem suite and the simple feed-forward neural network training task. We also show that the measures are negatively correlated with algorithm performance and therefore show potential as predictive meta-features for algorithm selection and other problem analysis. We envisage the proposed measures as complementary to other problem characterisation approaches, but with the advantages of not requiring any sampling of the search space, being invariant to transformations, and being very quick to calculate automatically.

Imaging device downtime is a major barrier to healthcare delivery in low- and middle-income countries (LMICs), often driven by limited access to specialized biomedical engineering support. We present a multi-modality medical equipment maintenance question-answering (QA) framework and demonstrate the fine-tuning of a medical foundation model for specialized technical troubleshooting tasks. Guided by a multi-country survey across nine LMICs, we curated technical manuals from MRI and ultrasound systems to generate the INGENZI_DatasetV1, containing 10,294 high-quality, filtered QA-context pairs. Using QLoRA-based parameter-efficient fine-tuning, we adapted the MedGemma-4b-it model to interpret system error logs and generate step-by-step equipment repair instructions. Compared to the baseline model, the fine-tuned system achieved substantial improvements across metrics, including F1 score (0.22 to 0.38), ROUGE-2 (0.18 to 0.41), and BERTScore F1 (0.86 to 0.91). These metric gains demonstrate that the model generates significantly more precise and procedurally accurate technical responses to new troubleshooting queries. This work establishes a reliable foundation for AI-assisted diagnostic and maintenance tools in resource-constrained settings.

Recommendation systems thrive on personalization, where ''correctness'' is rarely a binary truth but a matter of subjective human preference. As Large Language Models (LLMs) are deployed as autonomous verifiers of safety and quality guidelines, they face a distinctive challenge: context-aware preference alignment. Recent gains in Reinforcement Learning with Verifiable Rewards (RLVR) are indexed mostly on objective, mathematical tasks. Through a large-scale study spanning both proprietary and open-source models on four real-world verification tasks from a production recommender platform, we ask whether explicit reasoning generalizes to subjective, human-centric industry rubrics. We expose a fundamental vulnerability: rigid, math-centric reasoning traces actively degrade verification, and applying standard RLVR triggers a phenomenon we term reasoning collapse, in which the policy abandons deliberation in favor of rapid heuristic guessing. We introduce a conditional length-penalized post-training algorithm that intertwines verification accuracy with bounded reasoning length, halting collapse and recovering performance. Finally, we show that a reasoning trace's efficacy is tightly coupled with its socio-linguistic framing: across 1500 synthesized personas, verification accuracy swings by nearly 0.38 macro-F1 depending solely on the adopted reasoning persona---evidence that much subjective-verification error is really reasoning-style mismatch. This observation motivates a mid-training architecture that routes reasoning through contextually aligned personas. This work offers both a scalable algorithmic patch and a long-term architectural blueprint for aligning reasoning models with real-world subjective constraints.

Autoregressive transformers compute along two axes: horizontally across generated tokens, and vertically through model depth. Dense attention gives each token broad horizontal access to the past, but the vertical feedback channel between decoding steps remains narrow: only the sampled token returns to the bottom of the stack, while the top-layer hidden state is discarded. We introduce the \emph{full-bandwidth transformer}, which widens this channel with \emph{latent feedback}: at each decoding step, the previous top-layer hidden state is fused with the sampled token embedding through a gated linear unit and fed back as the next input. Latent feedback lets non-verbalized computation re-enter the stack with a renewed depth budget, while preserving the standard transformer architecture, KV cache, and language-modeling objective. To train full-bandwidth transformers without losing parallel teacher forcing, we use a scheduled multi-pass objective that introduces latent feedback late in pretraining and mixes a small fraction of deeper feedback passes for stability. We train 1B-parameter full-bandwidth transformers up to 400B tokens and find that latent feedback improves validation loss, 5-shot language-model evaluation, math and coding generation, and instruction-tuned performance. With negligible per-token decoding overhead, full-bandwidth transformers match or approach standard transformers trained with roughly $1.5\times$ more tokens, and manage to produce shorter reasoning traces at equal or better accuracy.

Reggaeton is one of the most widely consumed music genres in the world, and its lyrics are commonly regarded as highly sexualized. This claim rests mostly on qualitative studies and on small-scale quantitative ones. This paper has two goals. First, we present a reproducible method that uses a large language model to quantify thematic content in song lyrics along several independent dimensions. The method is not restricted to sexual content. Second, we apply it to a corpus of 1,259 songs by 12 reggaeton artists released between 2002 and 2025. The analysis covers four topics: a dataset characterization, a per-artist comparison, an analysis of how the dimensions change over time, and a comparison between our sexual-explicitness score and Spotify's own explicit flag. We release the data collection code, the scoring prompt, and the corpus, so that other researchers can replicate the approach or apply it to their own lyrics datasets.

Recent advances in retrieval-augmented generation (RAG) and large language models (LLMs) enable researchers to integrate AI into scientific workflows. However, using proprietary commercial AI systems raises concerns about transparency, reproducibility and privacy, which are essential for scientific practices. To this end, AquiLLM was developed as an open-source modular RAG-LLM framework using open-weight models, designed to support research groups in capturing tacit knowledge. In this work, we present a series of architectural improvements and feature enhancements to AquiLLM, including local embedding and reranking, multimodal capabilities, OpenAI-compatible inference interfaces, user interface improvements, semantic and episodic memory capabilities, and skills support. These enhancements were informed by discussions with domain experts, including astrophysicists and environmental researchers, and represent a step toward AI systems more closely aligned with scientific research practices.

AI tools that help people judge online claims are usually evaluated while the tool is present. This paper asks a different question: after using such a tool, what can the user still do on their own? I call this epistemic transfer. It refers to the effect of prior AI-assisted verification on later unassisted performance on new claims. In this paper, I make three contributions. First, I distinguish epistemic transfer from nearby outcomes such as correction effects, trust, reliance, and human--AI team performance. Second, I introduce two simple quantities for studying it: the Epistemic Transfer Effect (ETE), which compares delayed unassisted performance across conditions, and Tool-Removal Cost (TRC), which measures the immediate drop in performance when the tool is taken away. Third, I turn these ideas into a practical evaluation protocol that can be used in online experiments or field studies. The protocol combines answer-first and evidence-first AI conditions with active-practice and no-practice controls, delayed tests on held-out claims, behavioral measures, and participant- and item-level analyses. Putting ETE and TRC together yields a diagnostic space that separates capability building, capability plus tool advantage, epistemic inertness or de-skilling, and verification on loan. The point is not that every AI tool must teach. The point is that when independent judgment matters, we should test not only whether a tool helps now, but also what it leaves behind.

The current work developed seven Retrieval-Augmented Generation (RAG) models based on leading deception theories and compared how deception judgments were made relative to baseline models. Across 700 statements drawn from five published deception datasets, four large language models (gpt-4o, claude-sonnet-4-6, ollama/llama3, deepseek-v4-flash), and two run-types (RAG vs. baseline), a total of 39,200 deception judgments were rendered. Detection accuracies were consistent with typical human accuracies and not statistically different across RAG (54.5%) and baseline models (54.6%). RAG-based models (57.0%) were less truth-biased than baseline models (59.7%), but the effect size was quite small. Theoretical perspective mattered little for accuracy yet mattered substantially for response bias, which ranged from highly lie-biased (the verifiability approach, 32.2%) to highly truth-biased (truth-default theory, 88.1%). Content effects and model effects further moderated the results. Theory-guided AI judgments are unreliable with current parameters, yet they might show promise with additional datasets, model testing, and theory-to-data matching.

User Interface (UI) programming is challenging due to the complex abstraction gap between code and graphical software representations. To bridge this gap, UI programming tools often rely on screen retrieval and clustering, which require accurate similarity measures based on overlapping features. However, computing feature-oriented similarity is difficult because screens with similar functionality often exhibit design variations. To address this, we propose FRAME (ReinForced UseR InterfAce Screen EMbedding with Graphical Structural ComprEhension), a multi-modal, neuro-symbolic embedding technique. FRAME constructs symbolic, graph-based representations of UI components to encode salient relationships and capture feature patterns across different screens. It leverages large vision-language models for visual and lexical encoding, alongside a novel UI-specific computational geometry algorithm that enables weighted embedding propagation. Across three benchmarks, FRAME outperforms strong baselines by up to 13% MRR in search and 7.6 percentage points in clustering accuracy. A comprehensive ablation study further confirms the benefit of each component, demonstrating FRAME's potential for enhancing automated UI design and testing tools.

Transformer-based large language models (LLMs) achieve strong performance across many tasks, but their Key-Value (KV) cache grows linearly with sequence length, creating a severe memory bottleneck for long-context inference. Existing heuristic eviction methods (e.g., H$_2$O and SnapKV) rely on static attention or positional signals that often fail to capture a token's future predictive influence. We propose DistillCache, a reinforcement learning framework that formulates KV-cache eviction as a sequential decision problem. DistillCache learns a lightweight policy network using rich internal model signals (attention statistics, value norms, entropy, and position) and trains it with REINFORCE via a per-step KL-divergence reward to preserve the full-cache output distribution. On a 7B-parameter instruction-tuned Transformer (Mistral-7B-Instruct-v0.3), DistillCache retains 94.2% of full-cache accuracy on LongBench at a 25% cache budget, outperforming both strong heuristic baselines (H$_2$O, SnapKV) by up to 2.7 absolute points and, under our re-implementations, concurrent RL-based methods (ForesightKV, RLKV) by up to 1.4 points on long-context tasks. On reasoning benchmarks, DistillCache is competitive with the best concurrent method and surpasses it under aggressive compression. It also delivers up to 2.1x full-cache throughput while maintaining competitive practical efficiency. These results highlight the effectiveness of learned, distribution-aware policies for memory-efficient long-context LLM inference.

A graph layout is normally a table of $N$ free coordinates. We optimise a function with a fixed number of parameters instead. This gives a drawing a sample complexity and an extensible domain. Force-directed algorithms remain the standard tools for graph drawing. The most accurate among them minimise stress in the Kamada-Kawai formulation by directly optimising the node coordinates, at a full objective cost of $O(N^2)$ in time and space. Here, we propose Fling (Field Layout via Implicit Neural Geometry), a small neural network mapping the distances of each node to a set of landmarks, positioning it in the plane by training on the layout energy. The full spring system then becomes tractable without its distance matrix, as rest lengths follow from a landmark bound in constant time per pair while a second network learns the majorisation sums from exact anchor rows, at $O(|\mathcal{A}|N)$ per step for $|\mathcal{A}|\ll N$ anchors. Unlike neural drawers that read the graph by message passing, we represent the drawing as a function of node features. An unseen node costs one forward pass, where sparse and low-rank majorisation remain transductive. As the unknowns are weights rather than coordinates, the energy only requires a small fraction of the nodes, and a field fitted that way outperforms PivotMDS, landmark MDS, and a kernel ridge trained on the same energy and features, when the task is fitting the energy of a graph from a sample of its nodes. In addition, the same parameterisation enables a stochastic pivot stress variant, an aesthetics-optimised variant carrying a neighbour-embedding energy with node-edge clearance and crossing terms on the same field, and conditioning on the weight between two energies gives a whole layout family from one run.

Facial Emotion Recognition (FER) is an important task that has significant implications across various fields such as biometrics, health, and human-computer interaction. Current Vision Transformer-based approaches display quadratic complexity $\mathcal{O}(N^2)$, with N being the input sequence length, making them cumbersome to deploy at the edge. In this paper, we hypothesize that the FER task does not necessarily require all facial information to correctly interpret emotional states, as specific regions such as the eyes, the mouth, and parts of the cheeks carry discriminative information that can be sufficient to recognize emotions. Based on this, we propose Sparse Attention to Emotion (SAE), a model that discards image tokens that have no added value to the emotional context, while preserving good accuracy and achieving a significant gain in computational cost. Surprisingly, even after suppressing 90\% of the image tokens, our model achieves competitive accuracy to state of the art methods at much lower cost, providing a lightweight Facial Emotion Recognition approach. Experimental results demonstrate that SAE achieves new state of the art results on the RAF-DB dataset while reducing the computational complexity by up to 90\%.

In this paper we develop quantitative approximation results for shallow neural networks constructed using a dictionary based on metaplectic operators. First, we extend the concept of Barron spaces by considering a symplectically motivated extension of the Fourier transform, known as the metaplectic transform. Then, after establishing embedding between metaplectic Barron spaces and Sobolev spaces we consider a neural metaplectic dictionary and we prove Monte-Carlo approximation bounds for metaplectic Barron functions using finite linear combinations of atoms of the dictionary. Finally, we validate the introduction of the neural metaplectic dictionary by devising a deep neural network architecture that uses as building blocks the atoms of the dictionary. We test it to approximate solutions of time-dependent Schrödinger equations, demonstrating better performance compared to classical phyisics informed neural networks architectures.

Large language models are increasingly used for ordinal classification, yet semantically equivalent changes to prompt organization can alter their predictions. We conduct systematic experiments to characterize positional bias from label order, demonstration order, and demonstration placement. First, we apply the three probes to ten frontier LLMs on a common ordinal-classification task; every model is sensitive to all three positional sources, showing that the problem is pervasive. Second, we vary eight prompt-, task-, and model-level factors across five datasets; accuracy and stability are often misaligned, and only lower scale cardinality consistently improves both. Third, we compare pointwise, pairwise, and listwise inference, alternative aggregation and debiasing methods, and joint configurations; the tested corrections do not provide a reliable remedy, while a comparison-based listwise formulation offers the best balance but transfers unevenly across models and bias sources. These findings show that positional robustness depends on the full system configuration rather than the model alone. Ordinal-classification systems should therefore be selected jointly for predictive performance and stability.

Many modern AI systems analyze conversational traces to infer aspects of human interaction and state, implicitly assuming that such information is recoverable from conversation. We study observability: whether a target is recoverable from conversational transcripts alone. Observability is difficult to assess because transcripts may provide only a partial view of many targets, and large-scale analysis requires model-based annotation, making true limits of the conversational signal hard to distinguish from annotator error. We therefore study clinical encounters, where patient-reported outcome measures (PROMs) provide an external anchor for patient state, and visits follow broadly structured patterns. We study observability of patient state and conversational phase structure using 439 real-world clinical encounter transcripts spanning 134 hours, including 245 ENT transcripts paired with 273 PROM surveys. We operationalize patient state using PROM scores for voice, cough, and swallowing; phase structure using conversational phase segmentation. To make these analyses credible at scale, we use a PHI-compliant GPT-5 deployment for transcript annotation and conduct 40 hours of manual validation, reducing the risk that apparent limits of observability simply reflect annotator error. Our core finding is an observability asymmetry: phase structure is observable and useful for characterizing clinical encounter organization, while patient state is only partially observable, even in a setting designed to elicit patient symptoms and experiences, cautioning against transcript-only inference of human state.

Wireless Body Area Networks (WBANs) generate multivariate physiological time series that are highly nonstationary and must often be processed under strict computational and memory constraints. A critical yet underexplored challenge in this setting is selecting an appropriate temporal receptive field, which serves as a strong inductive bias for anomaly detection models. Existing approaches typically rely on fixed temporal contexts, which can perform inconsistently across heterogeneous signal regimes and require dataset-specific tuning. We propose ORCA, an agentically controlled anomaly detection framework that dynamically adapts the temporal receptive field at inference time based on lightweight signal statistics. Rather than introducing additional trainable parameters or learned policies, ORCA employs a supervisory controller that autonomously selects among discrete temporal contexts, enabling state-dependent inductive bias adaptation without retraining. Across a custom WBAN dataset, ORCA achieves performance comparable to the strongest fixed-context baselines (AUROC = 0.99) while eliminating the need to tune temporal horizons in advance. We further evaluate ORCA on MIMIC-IV as a challenging out-of-distribution benchmark, observing conservative generalization behavior without performance collapse under heterogeneous clinical conditions. These results highlight adaptive temporal inductive bias control as a practical and robust design principle for anomaly detection in resource-constrained, nonstationary physiological time series.

Sparse Mixture-of-Experts (MoE) routers commonly use the same scores both to select experts and to weight their already-computed outputs. We study whether these two roles, dispatch and aggregation, should be coupled. On pretrained OLMoE-1B-7B, we keep selected Top-8 expert IDs, expert computation, and total selected router mass fixed and change only within-set aggregation. A structured oracle improves full-horizon cross-entropy by 0.0160 +/- 0.0039 across three seeds; the router's top-scored expert is the counterfactual-best vertex only 17.2% of the time, with router-utility Spearman 0.030. We therefore train Fixed-Dispatch Adaptive Aggregation (FDAA), a 301K-parameter post-compute head optimized directly with the language-modeling objective while freezing the backbone, router, and experts. On OLMoE, FDAA improves fresh WikiText-103 test by Delta CE = -0.1523 +/- 0.0031 across three seeds, and mixed-domain training gives robust gains on WikiText-103, C4, and held-out Penn Treebank under frozen confirmatory evaluation. We also replicate the fixed-dispatch audit on DeepSeek-V2-Lite, which uses Top-6 routed experts plus shared experts. Best-vertex headroom remains significant on WikiText and C4, while router Top1 identifies the best selected expert in only 12.5% and 16.7% of audited examples. In a one-seed mixed-domain replication, FDAA improves locked WikiText and PTB, while C4 is statistically neutral. These results support a cross-architecture distinction between expert selection and expert commitment.

AI agents can now solve problems, answer like subject experts, and generate long-form multimodal content. However, whether they can adapt a lesson to fit a specified learner, which education calls Pedagogical Content Knowledge (PCK), has not been benchmarked. To measure it, we introduce the Teaching Monster Challenge, the first instructional video generation benchmark to treat the learner persona as an explicit evaluation criterion. Each system is given a topic and a learner persona and must generate a complete instructional video. Every video is screened by an LLM-judge, ranked by crowd pairwise voting, and finalized by an expert panel. The first edition shows that today's systems handle the content well but are far weaker at presenting it and adapting it to the learner. The same process exposes a limit of automatic judging. The LLM-judge separates a clear low-performing tail but ranks the strongest systems poorly. The strongest systems receive nearly identical scores from the judge, so its ranking of them does not match human preference. Progress therefore requires not only better teaching systems but also better automatic judges, and we release the benchmark, rubric, and human judgments as a testbed for both.

Subword tokenizers represent many common words twice in space-using writing systems, once with a leading space and once without. The two entries have separate embeddings in models, so occurrences of one word are divided across rows that are trained independently, and the two forms need not even segment the string the same way: " together" may be a single entry while the same word without a preceding space is tokenized as "to|gether". Capitalization divides a word further, into as many as six forms. We introduce an alternative to standard whitespace conventions using an explicit word boundary marker, which prevents such duplication. Words are delimited by the boundary markers, and spaces between words are represented as pairs of such markers. Two shift codes do the same for title case and upper case, allowing one internal representation of a word to be re-used across different settings. Switching to this convention mitigates the duplicate-entry issue, but does not improve tokenization compression: for both vocabulary-learning algorithms, the best marker scheme stays within one percent of the baseline in characters per token, averaged across six languages. It does result in better language modeling performance. Every marker scheme tested downstream reaches lower bits per byte than the baseline, suggesting that duplication carries a cost that compression does not capture.

Legal Statute Prediction (LSP) involves automatically identifying relevant legal statutes given factual descriptions in legal documents, typically framed as a multi-label classification task within natural language processing and information retrieval research. While recent advances have begun incorporating Large Language Models (LLMs) for statute prediction, current approaches primarily focus on accuracy metrics without addressing the critical need for legal reasoning, a fundamental requirement in judicial contexts where decisions must be explainable and justifiable. To address this research gap, we present PROSLEX (PRediction Of Statutes and LEgal eXplanation), a comprehensive dataset comprising 1,623 expert-annotated legal documents from the Indian context. Each document is paired with statute predictions and detailed explanations, totaling 7,450 explanations, capturing the underlying legal reasoning. Using this dataset, we systematically evaluate various prompting strategies, including zero-shot, few-shot, chain-of-thought, and tree-of-thoughts approaches, to generate both statute predictions and their corresponding legal rationales. Our evaluation framework measures not only predictive performance but also the coherence and legal validity of generated explanations, positioning PROSLEX as a benchmark for developing explainable AI systems that can support legal practitioners while advancing research in interpretable legal NLP. To ensure reproducibility, we have made our PROSLEX dataset and model code available on GitHub: https://github.com/subinay494/Legal_Statute_Prediction_Explanation.

Activation steering edits the behaviour of a frozen language model by adding a learned vector to its residual stream, and current practice fixes the injection layers globally per task. We argue that the best layers are an instance-level decision, and we make per-instance, multi-layer selection both well understood and deployable. On two open-weight 8B models and six binary persona traits, a per-instance oracle over layer subsets shows that the best layers vary from one input to the next: on most trait-model pairs, no fixed global layer set recovers the per-instance benefit. A greedy rule that ranks layers by single-layer marginal effect recovers nearly all of the oracle's benefit, but both must score candidate layers against the gold answer, so neither can run at deployment; the rule instead becomes the target a prompt-only predictor is trained to reproduce. Our deployable recipe needs no label at inference: a per-instance layer ranker read off the prompt embedding, a classifier that infers the steering direction, and an adaptive gate that scores short steered passes against that inferred direction and steers no more layers than necessary. The recipe recovers most of the oracle's lift (the bulk on the stronger model, a clear majority on the harder one), never drives any trait-model pair below its unsteered alignment baseline on average, and largely avoids the fluency collapse that strong global selection incurs at higher layer counts. A mechanistic account, "direction over magnitude", explains the behavioural flip under a mis-directed global set, the output collapse from steering too many layers, and the ceiling of unsteerable inputs.

Adaptive procedures must work without nuisance information an oracle may use, such as a gradient scale or smoothness index, and robust procedures may have to answer queries whose coordinate and inspection time are chosen only after the data are seen. Such comparisons are meaningful only when the oracle advantage and validity contract are stated explicitly. We formalize nuisance adaptation via a slice-normalized minimax ratio retaining the worst-case instance within each nuisance slice, and separately define the robustness cost of expanding from one preannounced Gaussian query to arbitrary post-hoc inspection. Our main result is a finite-horizon composition law for Gaussian certification: from M independent coordinates, a familywise certifier protecting every coordinate and time up to T pays optimal normalized squared half-width of order log(eM) + log log(e^eT), within the sample-mean-centered rectangular class. Epoch stitching gives the upper bound; independent Gaussian block increments across coordinates and geometric time scales give a matching lower bound, already holding on a geometric checkpoint grid, forcing quantiles of the realized maximum width so selection and stopping taxes add. Two benchmark regimes complete the picture: unknown gradient scale in online convex optimization has constant cost, while pointwise adaptation over nested Holder classes costs order (log n / log log n)^(s1/(2s1+1)). Cast as model monitoring, the law lets an analyst inspect any of M slice metrics at any data-dependent time: the naive fixed-query band's selected coverage degrades sharply, to 0.30 at M=1 and to zero for M>=10, while the epoch-stitched certifier holds familywise coverage at an additive iterated-logarithm width cost. Experiments put both sharp predictions at risk of refutation; both survive.

Foundation models for time series forecasting demonstrate impressive zero-shot generalization but often underperform on specialized domains such as high-frequency finance. We present a comprehensive study of hybrid neural-classical correction for adapting frozen TimesFM (200M parameters) to stock return prediction during the volatile opening trading hour. We compare two neural correction architectures - AttnCorrect (multi-head self-attention, approximately 471K parameters) and GatedLinear (low-rank bilinear projection with gating, approximately 49K parameters) - each augmented with Random Forest residual learning. Through systematic ablation across 10 major technology stocks (NVDA, MSFT, AAPL, GOOG, GOOGL, AMZN, META, AVGO, TSLA, NFLX) spanning 2 million data points, we reveal critical insights: (1) The hybrid neural-classical approach achieves 0.597 pooled correlation and 6.4x mean per-day correlation improvement over frozen TimesFM; (2) Classical residual learning (Random Forest) provides the largest single-component contribution, matching or exceeding the neural correction component; (3) Simpler neural architectures surprisingly outperform complex ones when classical residual learning is removed; (4) Self-attention provides the largest neural-only contribution. GatedLinear+RF achieves best overall performance with 9x fewer neural parameters than AttnCorrect+RF. We report three complementary correlation metrics - mean per-day, cross-day cumulative, and pooled - to provide a complete picture of predictive quality. Our results provide practical guidance: effective foundation model adaptation requires careful integration of neural and classical components, with classical methods playing a crucial complementary role.

Cognitive decision-making research depends on diverse scenarios with carefully controlled complexity, yet manual production is slow, inconsistent, and biased. We developed an automated pipeline that uses LLms to generate structured decision scenarios and validates their complexity through a composite framework rooted in established task-complexity theory. We evaluated 4,238 scenarios across multiple domains and complexity tiers. Measurement validation met rigorous psychometric standards. Agreement among five independent model families was nearly perfect, with an intraclass correlation coefficient of 0.997 and a kappa of 0.971. Known-groups validity demonstrated large separation between tiers, with an eta-squared of 0.587 and all pairwise comparisons significant at p less than .001. Factor analysis revealed a dominant complexity construct, with loadings between 0.87 and 0.96 across three frameworks, while interactivity formed a weaker secondary dimension at 0.34. Discriminant validity was limited by a strong relationship between complexity and text length that persisted after controlling for tier, yielding a partial correlation of 0.86. This constrains construct purity but does not undermine the instrument's tier-grading function. Model analyses showed a negative association between throughput and schema pass rate (r = -0.967, p = .007, n = 5), suggesting a speed-quality trade-off, though largely driven by one high-throughput model. Llama 4 Maverick generated scenarios fastest at 134 per minute versus 25 for DeepSeek Chat V3.2, but underproduced complex-tier scenarios, whereas DeepSeek Chat V3.2 balanced domain coverage with high schema compliance. The system demonstrated strong psychometric properties, enabling reliable classification into Simple, Moderate, and Complex tiers and providing the measurement infrastructure needed for downstream cognitive assessment of AI systems

Traffic sign recognition (TSR) models based on deep neural networks achieve strong clean-data performance but remain vulnerable to physically realizable adversarial attacks, including shadow perturbations, natural-light interference, and printed patches. Existing defenses often improve robustness against one attack type while degrading performance on others, and can reduce clean accuracy. We propose LAMDA (Language-Anchored Model for Direction Alignment), a training framework that transfers language-grounded structure into TSR models without using adversarial examples or adding inference-time overhead. LAMDA builds two fixed prototype banks from VLM-generated sign descriptions and class names using a frozen OpenCLIP text encoder, and uses them to supervise visual features through two complementary auxiliary losses during training. At inference, the adapter and prototype banks are discarded, leaving a standard backbone and classifier. Evaluated on GTSRB and LISA across four backbones and three physical attack types, LAMDA is the only method among ten evaluated that consistently improves robustness across all attack-backbone-dataset combinations, with gains of up to +12.5 pp under shadow attacks and +13.2 pp under natural-light attacks, while preserving or improving clean accuracy in nearly all cases.

We present 360CityArena, a benchmark for evaluating the urban exploration capabilities of embodied agents within a photorealistic environment constructed from 360-degree videos. Existing outdoor benchmarks either lack sufficient photorealism or complexity, resulting in a considerable gap from real-world urban environments. 360CityArena is built on a realistic reconstruction of the Akihabara district in Tokyo, Japan, using 602 360-degree video segments covering 85 streets, and consists of 175 meticulously human-crafted tasks. It encompasses three task categories: Environment Understanding, Path Reasoning, and Spatial Reasoning, covering fundamental abilities required for urban exploration, such as localization, landmark search, path planning, and relational spatial reasoning, thereby enabling comprehensive evaluation in realistic urban scenes. Our evaluation using state-of-the-art LMM-based agents shows that even the strongest model, Gemini 2.5 Flash, performs far below human level (human: 77.3% vs. Gemini 2.5 Flash: 17.1%), revealing substantial challenges that remain in city-scale embodied navigation and reasoning. 360CityArena provides a necessary and challenging testbed for photorealistic urban-district navigation and spatial reasoning.

A single model scale challenges the flexibility of a production retrieval system: some settings need it faster, others need a smaller index, and the right trade-off changes with the workload. In the context of information retrieval (IR), a transformer-based model can be made smaller in three ways---using fewer layers, passing fewer tokens through the upper layers, or producing a shorter embedding---and each way saves a different compute resource. These options have been studied one at a time, each as its own method with its own code and training setup, which makes them hard to combine or adapt to a new model. We present~\ours to bring all three under one simple abstraction: a single object names any size the model can run at, and a short schedule lists the sizes to train. Training then produces one checkpoint that serves all of those sizes, and at deployment the user picks any of them. The same abstraction covers both retrievers and rerankers and both encoder and decoder models, as it works through interfaces that Hugging Face transformers already expose; a new backbone is a configuration change, not new modeling code. Prior methods---Matryoshka embeddings, early exit, 2D~Matryoshka (e.g., Starbucks), and layerwise token compression---become special cases of our unified abstraction. The same interface also enables Matryoshka~LTC (MLTC), which jointly trains several token-compression ratios in one retriever checkpoint. To validate our framework, we train 20 checkpoints across three backbones and two tasks: the quality curves are smooth, one checkpoint costs little over a model trained for a single size, and a controlled study confirms the wallclock speedups. We release the framework and all checkpoints as a resource for building elastic retrieval systems.

Objective. Healthcare IT is usually organized by the technologies it adopts. We instead organize it by the unit of information a system makes computable, and describe a computational layer whose object is patient-specific clinical intent. Approach. We give criteria for a computational layer, derive three (record, clinical state, and a proposed layer of intent), and formalize the Actionable Clinical Record (ACR) as the atomic object of the third layer. Discussion. The framework distinguishes prescribed, observed, and intended process; existing standards represent intent once it is structured but do not recover it from natural communication, the capability we localize. The ACR is complementary to FHIR workflow resources, guidelines, and process mining; a companion feasibility study illustrates tractability for one narrow subproblem. Conclusion. Computable clinical intent is a coherent research direction; the ACR, its readiness ladder, and an executable-correctness evaluation framework are reusable constructs for subsequent work to extend, evaluate, or falsify.

With advancements in long-distance wireless power transfer (WPT) and space-based energy technologies, integrating WPT into non-terrestrial networks (NTNs), referred to as NTN-WPT, is emerging as a promising approach for next-generation wireless networks. This paper proposes an energy-scheduling approach that jointly optimizes energy efficiency, task completion rate, and task waiting time for power transfer from low Earth orbit satellites to terrestrial mobile user devices (UDs). To address scheduling challenges caused by satellite and UD mobility and channel uncertainty from stochastic propagation effects, we decompose the problem into three subproblems within a three-layer predictive framework: 1) a state prediction layer forecasts UD and satellite states; 2) an interaction mapping layer uses a graph neural network (GNN) to model energy transfer efficiency; and 3) a decision-making layer determines the energy allocation plan. Distinct machine learning (ML) methods are tailored to each layer. To balance the competing objectives, we adopt a multi-objective reinforcement learning (MORL) technique that scalarizes them into a weighted-sum reward, transforming the multi-objective problem into a tractable single-objective problem. We further introduce a multi-agent deep learning model integrating self-attention with multi-agent proximal policy optimization (MAPPO) to improve objective balancing. Simulation results show that the proposed approach achieves a better overall trade-off than baseline methods, maintaining competitive task completion rates and energy efficiency while reducing task waiting times, and remains robust under highly variable conditions.

Reinforcement Learning with Verifiable Rewards (RLVR) makes Multimodal Large Language Models more accurate, but the gains are brittle: simply paraphrasing a question or changing the prompt template can degrade them, which challenges reliable deployment in high-stakes scenarios like medical VQA. We trace this to two issues of the standard RL objective. First, the binary verifier conflates format with content, so the reward signal cannot tell a wrong answer apart from a misformatted one. Second, the training distribution covers only a thin slice of the real-world prompts that the model might meet at deployment, so policies that perform well on the training distribution can behave differently under unseen prompts during test. Both failures call for a robust post-training method that helps the policy cover a broader distribution of semantically equivalent prompts, and we identify two measures that help achieve this objective: separating format from semantics in the reward, and applying policy invariance across perturbed prompts with equivalent semantics. We therefore propose Prompt-Invariant RLVR (PIRL), consisting of a dynamic trinary reward and a consistency regularizer based on an embedding-space adversary. Under stress testing, PIRL's average accuracy on benchmarks drops by only $\le 1\%$, where GRPO drops ~3%. On dynamic evaluation, PIRL also achieves the smallest performance drop.

Translating technical requirements across languages can introduce semantic drift, altering numerical constraints, polarities, modalities, or other specification-critical meaning. IDRAAK is presented as an interpretable framework for detecting such drift using a language-independent Semantic Requirement Representation (SRR), with six detection workflows evaluated, ranging from deterministic comparison to multi-agent verification and few-shot prompting. On 890 synthetic perturbations across 300 requirements from 10 engineering domains, a single LLM call with six few-shot examples achieves MCC=0.888 and F1=0.983, outperforming the evaluated structured and multi-stage alternatives. Further evaluation on PAWS-X (805 pairs, 5 languages) and XNLI (700 pairs, 7 languages) exposes complementary strengths and limitations of structured and LLM-based approaches. Deterministic SRR comparison performs strongly on technical requirements (F1=0.898) but poorly on general-domain text (F1=0.012), while structured evidence improves performance on adversarial paraphrases. Post-hoc Platt scaling further improves confidence calibration. The results demonstrate that increased agentic complexity does not necessarily improve semantic-drift detection and that simple few-shot prompting can provide a strong and efficient alternative.

Pretraining LLMs on artificial languages ("pre-pretraining") is a technique that could reportedly increase token efficiency by 33%, i.e., save up to 33% of training tokens needed to reach a certain performance. We validate this prior result for English on a larger set of natural languages across four language families, using two different tokenizers and varying model sizes. We also relate the observed gains (or losses) in token efficiency to quantified linguistic properties of the languages, such as sentence length, morphological richness, and features of dependency syntactic trees (tree depth, number of children, number of crossing dependencies). Our empirical results indicate that the reported gains depend heavily on the experiment setup and the choice of random seed, although we can confirm the trend of stable gains with 128-Dyck pretraining of small models with the Llama tokenizer for most of the examined languages. On a general note, we argue that multiple training runs should be carried out at least for a subset of experiments to avoid the community adopting unstable approaches.

Tool-using large language model (LLM) agents are vulnerable to indirect prompt injection (IPI), in which malicious instructions embedded in external observations manipulate subsequent agent decisions and actions. Most existing adaptive attacks rely on repeatedly querying and refining against the target agent, whereas realistic attackers may have only a single opportunity to interact with an unknown target agent. We propose SAVOR (Strategy Abstraction Via Outcome-Conditioned Reflection), which shifts attack adaptation from test-time iteration to offline strategy distillation. SAVOR performs outcome-conditioned reflection over successful and failed trajectories collected from disjoint training environments, validates context-conditioned candidate strategies, and iteratively consolidates them into a reusable strategy memory. At test time, the frozen memory guides the generation of a single payload for each unseen target, requiring only one target-agent query and no target-agent feedback. Across two benchmarks and three victim models, SAVOR attains the highest average attack success rate in all six settings, leading the strongest prior attack by 2.5 to 11.8 points and the same injection channel without strategy learning by 23.1 points on Agent Security Bench, which holds out attacker tools, and 28.6 points on OpenClaw-IPI, an executable benchmark we introduce that holds out attack goals and verifies attacks through tool interactions and execution receipts. A memory learned under one defense also transfers to another.

Omni-modal LLMs jointly process audio, video, and text, but long multimodal sequences incur substantial prefill and KV-cache costs. Existing omni-modal compression methods primarily focus on pre-LLM token reduction, leaving modality-specific compression across the LLM boundary underexplored. We propose A-PACK, a two-stage framework that defers audio pruning until query-conditioned multimodal interactions emerge. Our analysis shows that audio exhibits higher task-relevant information density and representational diversity per token than video. We further find that local audio-visual dynamics provide a more effective cue for visual selection than token-wise matching. We therefore preserve audio and compress video with local dynamics before the LLM, then progressively prune low-relevance audio and visual tokens and their KV-cache entries inside the LLM. Across four benchmarks on Qwen2.5-Omni-7B/3B, A-PACK achieves the strongest average performance among the evaluated prior methods while reducing prefill FLOPs by up to 78% and improving decoding throughput by up to 2.21x.

Agent Skills package reusable instructions and assets for tool-using language-model agents. Progressive loading creates failure boundaries poorly represented by session-, model-, or tool-centric traces: a Skill can be discovered but not activated, activated without instructions, or appear successful without an independently verified outcome. We present Skill Runtime Intelligence, a passive runtime-intelligence system that reconstructs supported Skill-lifecycle stages across heterogeneous harnesses while preserving unsupported stages as unknown. Its Run Panorama separates immutable events, deterministic relations, inferred diagnoses, and controlled outcomes with four evidence grades; optional trace import and OTLP/HTTP export support existing observability deployments. Across six frozen repository profiles, three coding agents, and seven clean or fault-injected conditions, all 126 executions preserve source worktrees and each correlates to exactly one source session. Yet adapters expose three distinct semantics: no Skill runs; complete runs but no failure-like events; or failure-like events in every operational-failure and clean session. In a seven-template diagnostic study, semantic aliases and Panorama localize the same six non-clean boundaries but differ in exact/status behavior; both Raw views emit a failure status on all 18 clean cases, while Panorama emits none. A known-rule graph conforms to 126/126 frozen contracts, whereas a second model completes only 228/378 calls. These observations motivate executable adapter qualification and show that event presence is not boundary fidelity, composite exact scores mask distinct errors, and model explanations must not overwrite deterministic facts.

Diffusion language models use broad context to create text, suggesting they might handle input noise better than standard models. Testing reveals this is only partially true. Internally, diffusion models detect text errors highly accurately. Externally, their reported certainty ignores this signal. As accuracy drops due to noise, confidence stays near its maximum and the ability to correctly rank answers degrades toward random chance. We call this mismatch the representation confidence gap. The visible concentration of high certainty scores is a misleading surface symptom. Standard math adjustments remove this concentration but fail to fix the underlying loss of ranking order. This ranking deficit favors standard models under noisy conditions and resists common remedies. Matching training recovers accuracy but not ranking, while score recalibration and input level error signals cannot reorder the final answers. However, the information needed to properly evaluate an answer survives in the hidden states. A lightweight extraction tool uses this signal to improve ranking. This approach is highly efficient because it leaves the base model completely frozen and requires zero additional text generation steps. We present this tool to prove the signal exists, while clearly noting its limits. Ultimately, certainty reliability is a more pressing limit than overall accuracy under noisy conditions.

Koopman theory offers a linear-operator view of nonlinear sequence dynamics by lifting observations into a space where evolution is governed by a linear time-invariant Koopman operator. While the Koopman operator provides a linear representation of nonlinear dynamics, it is generally infinite dimensional and defined under time-invariant assumptions. To model non-stationary signals with frequency-dependent behavior, a frequency-varying extension is required. In recent years, deep learning has been increasingly employed to exploit its powerful function-approximation ability for learning the Koopman operator. In this study, we propose a novel approach called neural decomposition Koopman (NDKoop), an end-to-end architecture that integrates a learnable signal decomposition module with both frequency-independent and frequency-dependent Koopman based networks for sequence forecasting. To the best of our knowledge, this is the first work to jointly realize end-to end Koopman modeling and signal decomposition within a unified neural framework. We demonstrate that decomposing a signal into a frequency-independent trend component and a frequency-dependent periodic component, each governed by a corresponding Koopman operator, improves prediction accuracy when perfect linearization is unattainable. Numerical experiments across several forecasting benchmarks indicate that the proposed NDKoop provides strong performance.

Large language models (LLMs) increasingly serve as data-driven reasoners, yet their chains-of-thought (CoT) can be unfaithful even when final answers are correct. Most existing ``verification'' signals are not diagnostic: answer matching observes only the outcome, LLM-as-judge provides subjective and non-verifiable critiques, and scalar rewards (e.g., PRMs/RMs) offer little insight into where a multi-step derivation fails.We propose \textbf{SymDiag}, a neuro-symbolic framework that \textbf{reframes reasoning verification as structured failure diagnosis}. SymDiag translates natural-language CoT into symbolic constraints and performs step-level satisfiability/entailment checks to (i) localize failing steps and (ii) produce verifiable diagnostic evidence, including counterexamples, inconsistency witnesses, and missing-premise indicators. A central challenge is that apparent ``logic violations'' can be caused either by genuine reasoning defects or by neural-to-symbolic translation noise. SymDiag therefore incorporates a Self-Auditor that disentangles TranslationError from ReasoningError via dual symbolic encodings consistency checks, enabling robust diagnosis under partial observability. Across diverse mathematical, logical, scientific, and general reasoning benchmarks, SymDiag improves detection of unfaithful reasoning and provides substantially more effective feedback for multi-round reasoning repair than outcome-only verification and LLM-based judging, offering a principled foundation for trustworthy and scalable reasoning diagnosis.

In this study, we propose a quantum-classical physics-informed Kolmogorov-Arnold network (QCPIKAN) dedicated to the solution of fuzzy differential equations. The network takes the spatiotemporal coordinates and membership level as joint inputs and employs ChebyKAN modules and a parameterized quantum circuit to construct a hybrid function approximator. It simultaneously approximates the lower and upper endpoint functions associated with the α-cuts and incorporates the governing equations, initial-boundary conditions, and fuzzy-structural constraints into the training objective. Theoretically, a unified error-analysis framework is established for QCPIKAN and PIKAN, in which the endpoint-solution error is decomposed into approximation, sampling, optimization, and fuzzy-structure constraint errors. Under the assumptions of well-posedness and residual stability, it is proved that QCPIKAN has a smaller a priori error bound when the representational gain introduced by quantum entanglement features exceeds the additional computational error. Numerical experiments are conducted for elliptic, parabolic, and hyperbolic equations in an ideal quantum-simulation environment. The results show that QCPIKAN captures the overall contraction of the solution interval as increases. At most tested membership levels, the mean relative L2 error of PIKAN is approximately 1.1-2.7 times that of QCPIKAN. In the fuzzy convection example, the mean wavefront-position error of PIKAN is approximately 1.77 times that of QCPIKAN. Nevertheless, both models still exhibit local fuzzy-structure violations near boundaries, in high-gradient regions, and around the wavefront. These results indicate that QCPIKAN provides a quantum-classical hybrid physics-informed computational framework with comparatively high predictive accuracy for solving fuzzy partial differential equations represented by α-cuts.

Agentic benchmarks aim to measure how well AI agents plan, search, execute, and recover within realistic multi-tool environments, but they are almost exclusively in English. As AI agents are globally deployed to a linguistically diverse user base, whether agentic competence measured in English transfers to other languages remains an open question. We introduce OmnilingualGAIA2, a machine-translated expansion (with partial human- expert validation) of the GAIA2 agentic benchmark, covering ten target languages spanning five writing systems, paired with a localised and human-calibrated multilingual verifier. Evaluating seven frontier and open-weight agents, we find a universal cross-lingual gap of 8.8-18.4 pass@3 points that is agent-asymmetric in magnitude, concentrates on tool-orchestration rather than quantitative reasoning, and does not close with model scale. A stratified error attribution decomposes the gap as predominantly model-driven (55%), with a bounded translation-contamination floor of only 6.4% of scenario-language pairs. Human-expert linguistic analysis further identifies morphological cue loss and amplified ambiguity as the primary failure mechanisms in non-Latin-script languages. Our results argue that multilingual agentic evaluation must become a standard part of the reporting protocol for globally deployed agents.

Emotion is central to human communication, and its expression varies across languages. Large audio-language models (LALMs) achieve strong performance on multilingual speech tasks, yet it remains unclear whether they encode emotion through language-specific correlations or language-agnostic representations. We present the first neuron-level interpretability study of this question. We define Multilingual Emotion Neurons (MLENs) as functional units exhibiting stable emotional selectivity and aligned causal effects across languages, and introduce Consistency-Regularized Fusion (CR-Fusion) to identify them. Across four modern LALMs and 12 typologically diverse languages, emotion-sensitive neurons identified independently per language show minimal overlap, and additional monolingual identification data saturates quickly without isolating more transferable units, motivating identification from pooled cross-lingual evidence. Causal interventions demonstrate that MLENs identified by CR-Fusion provide more precise and transferable affective control than monolingual neuron sets in both zero-shot and low-resource settings. Leave-one-out ablations further reveal asymmetric transfer: individual identification languages, including low-resource ones, contribute non-redundant evidence, while several low-resource languages benefit most from the resulting cross-lingual transfer. Together, our findings provide the first causal, neuron-level account of how LALMs encode emotion across languages, and establish multilingual neuron identification as an effective mechanism for understanding cross-lingual affective behavior.

Diffusion models are increasingly used as controllable samplers, whose generations can be steered at inference time according to a chosen reward function. While such rewards are typically defined on individual samples, for many applications it is desirable to steer according to distribution-level rewards, for example to calibrate with population-level information or to encourage diversity. In both cases, simply incorporating the reward gradient into the dynamics, while often effective, comes with few theoretical guarantees on the sampled distribution. For pointwise rewards, recent work has therefore sought to develop a principled framework for targeting a prescribed tilted distribution using particle reweighting. However, an analogous theoretically-grounded approach for distributional rewards is currently lacking. In this work, we formulate inference-time distributional control as targeting a tilted measure under a mean-field framework, and derive a weighted interacting particle scheme to target it in a principled manner. Our framework recovers pointwise-reward steering as a special case, while providing a theoretical foundation for existing batch-level steering methods. Empirically, we verify that the procedure correctly targets the prescribed distribution in tractable low-dimensional settings, and investigate its behaviour in higher-dimensional protein conformation tasks.

On-policy self-distillation improves language-model reasoning by querying a teacher on states actually visited by the student. Recent methods create a powerful information asymmetry by exposing the teacher to privileged context, yet they fundamentally rely on external supervision---such as gold solutions or verifiers---to construct this advantage. We introduce CoDA (Consensus and Disagreement Alignment), a fully unsupervised framework that creates reliable privileged information entirely from the latent uncertainty structure of a model's own unlabeled rollouts. CoDA extracts two complementary signals. In the positive branch, answer-level consensus identifies a stable reasoning mode, which conditions a frozen self-teacher to provide dense distributional guidance on fresh student trajectories. However, because agreement does not guarantee correctness, positive-only distillation risks amplifying correlated errors into a false consensus. To break this harmful feedback loop, CoDA incorporates a negative branch that exploits disagreement: minority trajectories are treated as unstable alternatives and gently penalized via a reference-anchored, KTO-style calibration objective. This unpaired binary feedback provides robust regularization without requiring the strong assumption that the consensus is the absolute ground truth. Empirical evaluations on competition-level mathematical benchmarks demonstrate that CoDA significantly improves reasoning, outperforming self-generated baselines and effectively stabilizing training against erroneous consensus.

In an era defined by escalating climate change and the pervasive deployment of edge intelligence, the environmental cost of semiconductor manufacturing and operation has reached a critical threshold. As Deep Learning (DL) accelerators dominate System-on-Chip (SoC) die area, achieving true sustainability requires a paradigm shift from static worst-case efficiency to dynamic energy-proportionality. This paper introduces Eco-SoC, a highly scalable VLSI architecture co-designed specifically for sustainable artificial intelligence. We propose a hardware-level Dynamic Precision-Scaling Logic (DPSL) framework that adaptively modulates bit-width precision based on real-time activation sparsity, successfully reducing switching activity by up to 42% on a commercial 7nm FinFET process node. Furthermore, we transcend traditional Power-Performance-Area (PPA) metrics by providing a comprehensive Life Cycle Assessment (LCA) using the Architectural Carbon footprint Tool (ACT). Our synthesis demonstrates that Eco-SoC offsets its increased embodied carbon footprint (a marginal 4.8% area overhead) within 1.1 years of edge deployment. Finally, by introducing a thermal-aware power gating mechanism that mitigates localized hotspots, Eco-SoC doubles the projected Mean Time To Failure (MTTF) of the silicon, providing a tangible, scalable strategy for electronic waste (e-waste) mitigation in next-generation computing systems.

Prospective daily symptom tracking is central to premenstrual health assessment, but repeated ordinal forms impose substantial response burden. We formulate conversational administration as an ordinal label-recovery problem: the system actively elicits a small set of symptom clusters and maps each response to the original severity labels. We used 3,320 complete participant-days from the mcPHASES dataset, covering cramps, mood swing, fatigue, sleep issues, stress, and bloating on a six-level scale. Six participants were reserved for development and 36 for a frozen evaluation comprising 360 participant-days and 2,160 item labels. A ModernBERT evidence gate detected whether a symptom was expressed, and Qwen2.5-1.5B-Instruct produced deterministic structured severity scores. Fixed six-item questioning achieved a quadratic weighted kappa of 0.976, whereas three joint symptom-cluster questions achieved 0.913, 97.45% agreement within one severity level, and 80.94% recall for moderate-or-higher symptoms while reducing questions by 50%. Open-first adaptive policies required 3.92-5.98 questions and produced lower agreement than the corresponding fixed policies. Participant-cluster bootstrap analysis estimated a kappa difference of -0.062 (95% CI -0.076 to -0.048) between the three-cluster and six-item strategies. Active cluster-level elicitation provides a direct, local-model route from natural conversation to reusable daily symptom labels.

The carbon footprint of any deployed Large Language Model (LLM) accumulates during inference, where repeated use of the model substantially exceeds the one-time cost of fine-tuning. Yet most efficiency interventions target either pre-training scale or post-hoc compression. We ask whether folding a calibrated, differentiable energy surrogate into the fine-tuning objective can produce inference behavior that gains task accuracy at zero or near-zero carbon cost, a break-even configuration. We propose a joint loss mechanism with a per-model carbon-emission parameter, a linear surrogate over parameter norm, FLOP proxy, and a memory proxy, fit from on-hardware energy profiling. We fine-tune three architecturally distinct families: Gemma-2 2B, Llama-3.1 8B, and Qwen-2.5 14B, and evaluate inference F1 and CO$_2$ emissions on three MMLU subjects: abstract algebra, philosophy, and formal logic. We discover from several outcomes that the carbon term behaves as either harmful interference or beneficial regularization depending on the task structure. We position calibrated carbon-aware fine-tuning as a lightweight, drop-in regularizer with a non-empty but model and task-dependent break-even region. This is an ongoing work, and we will release our codebase soon.

Reinforcement learning (RL) seeks to optimize sequential decisions to maximize population-level benefits over time. However, when deployed in high-stakes settings such as healthcare, RL decisions might systematically restrict some subpopulation's access to valuable services in a manner contrary to the values and goals of stakeholders. Counterfactual fairness (CF) offers a promising framework to address this problem based on causal reasoning. This paper develops a data preprocessing algorithm that, when used in tandem with policy learning, enables CF in RL. Our algorithm relies on a novel quantile distribution mapping method for sequentially estimating the counterfactual states and rewards in the data preprocessing step, subsuming common additivity assumptions used for counterfactual prediction as a special case. We theoretically prove that the per-step level of counterfactual unfairness and infinite-horizon suboptimality gap can be bounded under mild regularity conditions. We also empirically test our algorithm in numerical experiments as well as in application to a real-world interventional digital health dataset.

Activation checkpointing minimizes the runtime of neural networks under a given memory budget, by selecting which intermediate tensors to store and which to recompute. PyTorch solves this as a 0/1 knapsack problem, where operations from a joint forward-backward computation graph are items with a memory cost (weight) and a runtime saving (value). The default solver, dp_knapsack, allocates a full dynamic programming (DP) table of shape $(n+1) \times (W+1)$, where $n$ is the number of operations and $W$ is the quantized memory budget. This method is resource-hungry and crashes at $n = 100$ items on a machine with 64 GB RAM. In this paper, we introduce dp_knapsack_sliding_hirschberg, which combines the sliding window trick and Hirschberg's algorithm to reduce peak memory from $O(nW)$ to $O(W)$ while preserving the exact optimal solution. Our experiments show successful knapsack execution at $n = 2000$, where dp_knapsack fails at $n = 100$, a 20$\times$ increase in computable problem size. In addition, our benchmarks show a consistent 25-28\% runtime speedup over dp_knapsack. The implementation is merged into PyTorch and released in version 2.10.

Fitness Action Quality Assessment (AQA) is important for intelligent sports training, yet the capabilities of Multimodal Large Language Models (MLLMs) in this setting remain underexplored. Existing benchmarks rely on action-specific annotation schemes and focus primarily on final assessment outputs, offering limited insight into how models assess exercise quality. We introduce FitAQA, a systematic benchmark for evaluating MLLMs in fitness AQA, containing 2,219 videos and 5,512 QA instances across 30 bodyweight exercises. In collaboration with experts in sports science, we develop a unified form error taxonomy that defines 38 recurring form errors within six complementary quality dimensions: alignment, symmetry, stability, coordination, tempo, and completeness. This taxonomy provides a shared assessment framework across different exercises. FitAQA further formulates three evaluation tasks: perception for recognizing relevant visual evidence, judgement for combining that evidence with domain knowledge to assess execution correctness, and temporal grounding for localizing form errors over time. Extensive evaluation shows that current MLLMs still struggle to assess exercise quality comprehensively and localize form errors precisely. Controlled experiments further indicate that visual perception is a key bottleneck, as judgement performance improves substantially when ground-truth perceptual evidence is provided. The dataset and evaluation code will be made publicly available.

Multi-vector vision-language retrievers enable fine-grained Visual Document Retrieval (VDR) through late interaction, but storing and scoring hundreds of visual patch embeddings per page incurs substantial overhead. Existing training-free methods rely on pruning or merging: pruning degrades sharply under aggressive compression, whereas merging does not explicitly prioritize important regions when forming representatives. We introduce AnchorFold, a training-free focus-then-fold framework for document-side index compression. AnchorFold applies Recursive Attention Propagation over visual self-attention graphs, performing multi-step propagation within each attention head and integrating scores across heads and layers. The focus stage selects the highest-centrality tokens as anchors. The fold stage assigns remaining tokens to their most similar anchors in the normalized retrieval space and summarizes each anchor-centered group through centrality-weighted aggregation. This preserves non-anchor contributions while concentrating capacity on structurally important tokens. Across ViDoRe v1/v2 and REAL-MM-RAG with three diverse retrieval backbones, AnchorFold consistently outperforms all evaluated training-free baselines at $γ\leq 0.20$. On ViDoRe v1/v2, it retains 98.3% of full-index NDCG@5 on average at $5\times$ compression, achieving near-lossless compression, and 92.4% at $20\times$ compression.

Large Language Models are deployed to multiple types of environments, from internet browsers to edge devices, and WebGPU serves as a modern cross-platform standard. The engines for browser-based LLM inference have proliferated, yet the overhead of WebGPU per-operation dispatch remains poorly characterized. In this work, we introduce a sequential-dispatch measurement method and show that naive single-operation measurements overestimate per-dispatch cost by conflating dispatch with synchronization. Using our method, we measure the per-dispatch cost and show that it is independent of data type used. We show that the dispatch overhead, not kernel quality, is the bottleneck at batch size 1, and isolate the dispatch count as the cause. Therefore, we conclude that at batch size 1, the effective approach to LLM inference optimization in WebGPU is reducing dispatch count. Our findings point to dispatch amortization, in the inference engines and in the WebGPU specification, as a path to practical browser-based inference.

To address this gap, we introduce TomaMMU, a large-scale Tomato leaf disease MultiModal Understanding dataset, alongside TomaBench, a benchmark for evaluating VLMs on tomato disease understanding. TomaMMU comprises 28,808 high-quality images spanning 15 categories and 213,119 human-annotated visual question-answer pairs, generated through a three-stage pipeline comprising Data Collection, Human Annotation, and Question-Answer Generation. Building on this foundation, TomaBench organizes seven agricultural tasks into a hierarchical three-level taxonomy spanning Basic Perception, Pathology Understanding, and Expert Diagnosis, which together enable systematic evaluation from low-level visual recognition to high-level diagnostic reasoning. The tasks assess visual symptom recognition, taxonomic relationships, and diagnostic reasoning, offering a comprehensive view of how well models grasp plant pathology. Our results pronounced gaps in fine-grained recognition and factually grounded reasoning with 14 state-of-the-art VLMs, consistently underperforming on both challenging MCQs and open-ended questions. These results suggest that current VLMs struggle to translate visual perception into reliable diagnostic knowledge, motivating the need for targeted domain adaptation. Simple fine-tuning on TomaMMU substantially narrows this gap, boosting accuracy on challenging MCQs to 96.09%, outperforming recent VLMs, and pointing toward promising directions for future work. All data and code is available in https://huggingface.co/datasets/enalis/TomaMMU.

On-policy self-distillation (OPSD) uses a privileged teacher to supervise a reasoning model on prefixes sampled from its own rollouts. Yet each rollout also reveals how the student's response unfolds and whether it succeeds, student-specific hindsight that standard OPSD does not use to form the teacher. We introduce Privileged Adaptation from Student Trajectories (PAST), which treats each completed student trajectory as additional privileged information for the OPSD teacher while leaving the student's distillation prefixes unchanged. PAST preserves the student's next-token distribution on correct trajectories and uses failed trajectories to adapt the teacher toward verified success under student-proximity regularization. We characterize what such a trajectory-conditioned teacher can transfer to a prefix-only student. Forward-KL distillation projects the teacher distributions to their conditional arithmetic mean given the prefix. This projection separates trajectory-specific variation that remains privileged from the mean policy shift available to the student. For correct trajectories, the unclipped population objective also has the frozen student as an ideal distributional fixed point. Across three mathematical reasoning benchmarks, PAST improves the Avg@12 macro average over Vanilla OPSD by 5.6 percentage points. A $2\times2$ factorial study shows gains from both complete-trajectory access and teacher adaptation, while trajectory removal and shuffling confirm that the adapted teacher uses the matching hindsight context.

Benchmarks for systems that are optimized against the evaluation signal measure something different from what they claim. We document this concretely in two GPU-kernel-optimization suites with held-out generalization gates: Metal-Sci (10 scientific-compute tasks) and Metal-ZK (12 zero-knowledge/cryptographic tasks), in which three frontier LLMs (Opus 4.7, Gemini 3.1 Pro, GPT-5.5) propose Metal kernels inside a $(1{+}1)$ evolutionary loop with rich feedback. Although no model is prompted to act adversarially, the promoted winners repeatedly fingerprint the evaluation configuration: they branch on the identity of runtime parameters, tune the measured branch maximally, and leave the unmeasured branch slow or silently wrong. Across the pooled suites, $16/53$ ($30\%$) of in-distribution wins fail to transfer to held-out configurations. We give a four-mode taxonomy of these failures, from configuration fingerprints to gate leakage. We distill design guidance for measurement under strategic optimization: held-out probes retain validity only on non-enumerable axes; gates must measure held-out performance, not just correctness; and a transfer rate is interpretable only with per-failure mechanism grades: ours decomposes into gamed, overfit, and benign. Code and research artifacts: https://github.com/vicgalle/kernel-fingerprinting

Speculative decoding accelerates large language model inference by drafting multiple tokens for parallel verification, with efficiency critically determined by the speculative length selected at each decoding round. Existing dynamic speculation methods select the speculation length by estimating how many tokens will be accepted, which is reasonable for autoregressive drafters that generates tokens sequentially. The recent wave of diffusion-based drafters, however, generates candidate blocks in parallel at substantially lower drafting cost, shifting the key question from how many tokens to generate to how many generated tokens are worth verifying. We therefore reformulate dynamic speculative-length selection as expected-speedup optimization and derive a marginal criterion that extends the speculative sequence only when its acceptance gain outweighs the additional verification cost. Building on this criterion, we develop \textit{LibraSpec}, a training-free and plug-and-play algorithm that iteratively determines the speculative length using drafter confidence scores. Theoretically, we prove that LibraSpec monotonically converges toward the optimal speculative length. Experiments across six target models, three diffusion-based speculative decoding methods, and math, coding, and chat benchmarks show consistent improvements under both greedy and sampling settings, achieving a further $0.5\sim1.5\times$ improvement over baselines and up to $8.49\times$ speedup over autoregressive decoding.

Vision-language models offer a promising path toward automating radiology report generation, but applying them to full 3D CT volumes poses substantial computational challenges. Modern foundation vision encoders (VEs) can produce tens of thousands of vision tokens per scan, making the visual sequence passed to the large language model (LLM) a primary computational bottleneck. Vision-to-language projectors can compress this sequence to reduce computation, but may discard clinically relevant detail; conversely, effective compression can accommodate higher-resolution inputs while keeping the downstream token count fixed. How this vision-token budget should be allocated across input field of view, spatial resolution, and vision-to-language projection therefore remains an open design question. We systematically evaluate four heterogeneous VEs (CNN- and ViT-based), five token-reducing projectors at up to 64x compression alongside a non-reducing MLP projector baseline, and five instruction-tuned LLMs (1.7B--4B) on two large-scale CT report datasets (CT-RATE and Merlin). At matched LLM token budgets, anatomy-guided region of interest cropping is the most consistent strategy, improving clinical macro F1 in 19 of 20 settings by +3.7 points on average for the 3D ViT Primus encoder and +1.1 for the slice-based 2D ViT Curia encoder. Increasing input resolution further is strongly projector-dependent: the PerceiverResampler, paired with higher-resolution Curia features, yields the strongest configuration in the resolution study on both datasets. Our best configurations achieve state-of-the-art clinical macro F1 on the test sets, reaching 49.5 on CT-RATE and 49.0 on Merlin. Code and models will be published upon publication.

The reliability of AI generative models is typically measured by output correctness, yet in practice it depends on the effort required to verify those outputs. We argue that current evaluation metrics overlook a critical failure mode: Verification-Cost Errors (VCEs), defined as incorrect input-output pairs that a declared fraction of the verifier population fails to identify within the verification budget available in a given deployment context. Unlike standard notions of "hallucination", VCEs are defined operationally, by the failure of correct identification within budget rather than by any property of the output itself. Plausibility and authoritative presentation are hypothesised contributors to that failure, not defining conditions. To capture this asymmetry, we introduce the notion of verification cost relative to a deployment budget as an operational dimension that current evaluation does not routinely capture. The quantity is presented as a conceptual instrument rather than a finalized metric. Evidence from code generation and multi-modal document understanding shows that high benchmark accuracy can mask significant verification effort in practice. We therefore take the position that correctness alone is insufficient as a measure of reliability. AI evaluation should explicitly account for verification cost, reflecting whether errors can be detected under realistic resource constraints.

Kernel spectral clustering with a single bandwidth can be inadequate for data exhibiting multiple characteristic pairwise-distance scales, a problem particularly prevalent in the high-dimensional regime. We address this issue through a multi-kernel formulation that aggregates kernels with different bandwidths. The bandwidths are selected as prescribed empirical quantiles of the pairwise squared distances, thereby capturing the relevant distance scales without requiring prior population-scale information. We develop a rigorous theoretical analysis of the resulting method under a general high-dimensional, multi-scale mixture model with heterogeneous cluster centers and covariance geometries. We construct a blockwise constant, low-rank informative approximation to the empirical multi-kernel matrix and establish row-wise $\ell_{2,\infty}$ perturbation bounds for its leading spectral components, as well as for the associated normalized Laplacian matrix. These bounds yield observation-level control of the spectral embedding, which is more informative than conventional global eigenspace perturbation estimates. Under suitable eigen-gap and cluster-separation conditions, we show that approximate $K$-means applied to the multi-kernel spectral embedding achieves exact recovery with high probability.

Reliable evaluation of tool routing is critical as Large Language Models increasingly operate as autonomous agents. Current benchmarks face three structural limitations: data distributions that follow a power law leave rare scenarios underrepresented; the absence of adversarial hard negatives obscures performance differences across models; and annotation pipelines depend on LLM judgments that have not been validated through execution. In this paper, we introduce PluginEval, a benchmark constructed through a two-stage framework that systematically mitigates these limitations. First, we formulate tool routing as a sequence of three decisions and separate generation from verification. LLMs propose candidate calls, while deterministic validation and real API execution provide reliable quality signals. Second, we decompose each plugin by capability, intent, and boundary to identify trigger and exclusion scenarios. We then generate queries at different difficulty levels to fill coverage gaps, including adversarial negatives targeting three failure modes, and return them to the first stage for annotation. This process creates a closed loop that iterates until coverage converges. For evaluation, we move beyond aggregate accuracy. An LLM judge anchored to gold annotations classifies failures as missed calls, spurious calls, or parameter errors, producing a detailed error profile for each model. We evaluate five model families, including proprietary models and models with open weights, analyze their performance across difficulty levels and error categories, and validate the judge through agreement with human annotations.

Video token communication represents video content as discrete tokens that differ in their importance to reconstruction and exhibit temporal dependencies. When these tokens are packetized for wireless transmission, block fading can cause multiple important or correlated tokens to be lost together, severely degrading video reconstruction. To address this issue, we propose a loss-resilient wireless video token communication (WVTC) framework. WVTC evaluates token importance from the intrinsic predictive structure of video tokens, assigning high priority to structural I-tokens and measuring P-token importance by temporal neighborhood novelty. A shuffled mixed I/P-token packetization scheme disperses structural anchors and correlated temporal regions across packets. Using only current block channel state information, an online scheduler jointly considers packet importance density, MCS-dependent decoding reliability, block capacity, and importance concentration when allocating packets to fading blocks. At the receiver, a fine-tuned detokenizer reconstructs missing content from surviving tokens without retransmission. Numerical results demonstrate improved perceptual quality and more graceful degradation under increasing packet error rates.

Residential virtual power plants (VPPs) can provide grid flexibility by shifting household demand, but physical flexibility becomes dependable capacity only when residents authorize a plan and the promised response is delivered. Existing benchmarks evaluate control but omit event-specific authorization. We present EnergyBridge, a benchmark and agent framework connecting capacity reporting, household authorization, and physical execution. It combines region-specific EnergyPlus environments for Tianjin and Berlin with an LLM-based User Participation Simulator. Against 584 persona- and event-matched human role-play judgments, the LLM-based User Participation Simulator preserves method ordering with a 5.3-point mean absolute acceptance error. Across conventional controllers and agent baselines, EnergyBridge achieves the highest simulated authorization, lowest event-window energy, and the most reliable capacity commitment in both regions. We release human data and codes for reproducible human-centered grid-flexibility research: https://github.com/Agentic-Intelligence-Lab/EnergyBridge.

The state evolution of a complex system arises jointly from object laws, relational propagation, domain conservation, and unmodeled error. Forcing all sources into one black box makes mechanism attribution and constraint preservation unauditable; forcing every mechanism into one equation family discards mature domain solvers. We propose SD-GWM, a Structural Dynamics Graph World Model as an executable structural contract: nodes declare self-dynamics S, edges declare neighbor graph-coupled dynamics N---both fixed-form mechanism assets (rules, ODEs, solvers) calibrating only authorized parameters. An optional bounded residual R concentrates learnability, while a global projection maps states to feasibility, enforcing constraints without guaranteeing accuracy gains. On eight pre-registered research questions, SD-GWM delivers (i) heterogeneous integration: rules and solvers plug in natively; (ii) semantic fidelity: disabling R preserves source semantics bit-for-bit, with four theory properties under explicit proof/empirical boundaries; (iii) auditable governance: stepwise traces enable counterfactual fault localization (top-1 = 1.0) without post-hoc approximations. On a semi-synthetic flood testbed and USGS streamflow, SD-GWM reduces constraint violations to floating-point tolerance in analytical tests and to zero in semi-synthetic and real-data cases. Persistence matches SD-GWM in calm periods, but during a 254-day extreme-flood shift persistence and all neural baselines collapse (90-min RMSE 892-3007 cfs) while SD-GWM holds at 108 cfs (8-28x gain). The bounded residual cuts RMSE ~50% only under backbone bias. We position SD-GWM not as a universally superior forecaster, but as a verifiable substrate for auditable, constraint-safe spatiotemporal mining.

Long-context LLM inference is bottlenecked by KV cache memory, yet distributing a limited cache budget across layers remains challenging. Existing methods rely on proxies such as layer depth, attention statistics, or representation change. These proxies do not measure how perturbations at each layer propagate to the output and may therefore cause sensitive layers to be underallocated while tolerant layers are overallocated. To address this issue, we propose RippleKV, which allocates cache across layers by estimating how perturbations to each layer's value cache affect the final predictive distribution. RippleKV independently injects norm-adaptive perturbations into each layer's value cache and measures the induced KL divergence at the model output over a small calibration set. Averaging these responses yields a sensitivity profile specific to the model that need not vary monotonically with depth. RippleKV then converts the sensitivity profile into layer budget multipliers by normalizing the sensitivity scores and applying an exponential mapping. A ratio parameter controls the allocation disparity between sensitive and tolerant layers, while a final normalization preserves the KV cache budget. Experiments on LongBench demonstrate that RippleKV achieves the highest average performance among the evaluated KV cache compression methods under matched cache budgets.

Skill evolution improves agent skills through feedback over time, with failed trajectories often providing informative signals by revealing incomplete or misleading behaviors. However, existing methods mainly rely on single trajectories, where early reasoning errors can propagate through subsequent steps and weaken the feedback available for skill refinement. Consequently, improving skills requires repeated cycles of rollout, diagnosis, and update, incurring substantial token costs. To address this challenge, we introduce Branch2Skill, an efficient framework that transforms a single reasoning tree into dense supervision for skill evolution. For each task or problem, Branch2Skill performs Monte Carlo tree search under a fixed budget to obtain diverse reasoning trajectories, then compares an elite path with sibling alternatives sharing the same prefixes to extract step-wise evidence about which reasoning patterns to retain, revise, or avoid. Finally, Branch2Skill distills multi-step evidence into reusable updates, allowing one reasoning tree to provide supervision across multiple reasoning steps and reducing the need for repeated rollout-update cycles. Across six benchmarks covering reasoning and agentic tasks, Branch2Skill consistently improves task performance while enhancing skill evolution efficiency. For example, with GPT 5.5 as the target model, Branch2Skill uses 73.2% fewer tokens than SkillOpt, while achieving superior performance. These results demonstrate that reasoning trees can support not only more effective trajectory search, but also richer supervision for more efficient skill improvement. Code will be published.

Semantic vision encoders have become a central visual interface for multimodal understanding and semantic conditioning in image generation. However, their final tokens discard fine-grained visual details, leading to poor pixel reconstruction and limiting their use in reconstruction-sensitive tasks such as image generation and editing. In this work, we ask whether understanding, generation, and editing can be modeled in a single visual representation space built from a pretrained semantic ViT. We show that the frozen Transformer blocks of a semantic ViT are not intrinsically unable to preserve visual details. Instead, the original patch parameterization drives the representation toward semantic abstraction, making fine-grained information difficult to recover from the final tokens. Based on this observation, we introduce \emph{Patch Reparameterization}, which preserves the original semantic pathway while adding a reconstruction-aware patch embedding that provides fine-grained visual information to the same frozen ViT blocks. The resulting unified representation preserves multimodal understanding while enabling high-fidelity image reconstruction and a favorable reconstruction--generation trade-off. We further scale this representation into \emph{UniSpace}, an 8B Mixture-of-Transformer-Experts model that performs understanding, generation, and editing in the same visual space without a separate VAE pathway. System-level evaluations demonstrate practical text-to-image generation and instruction-based image editing, showing that a reparameterized pretrained ViT can serve as a unified visual interface for scalable multimodal modeling.

Long-term time series forecasting benefits from preserving global structure such as trends and seasonality. Recent LLM-based forecasters often improve accuracy through test-time scaling (e.g., iterative refinement), but these methods are computationally expensive and increasingly prone to global-shape mismatch as the prediction horizon extends. We propose SCALER, a coarse-to-fine forecasting framework that first employs a lightweight Transformer tailored to long-term shape modeling to predict a coarse representation of future dynamics. This predicted shape then serves as a compact guide for an LLM to perform test-time scaling via iterative coarse-to-fine residual token refinement, while processing substantially fewer tokens at each step. By guiding refinement with an explicit future-shape prediction, SCALER reduces reliance on long description prompts, and its fixed-step refinement avoids costly reward-model-based selection, further lowering computational overhead. Experimental results demonstrate that SCALER outperforms strong forecasting baselines in long-term, short-term and zero-shot forecasting while significantly reducing the inference cost associated with scaled LLM for time series forecasting. Code: https://github.com/xuanmay2701/SCALER.

In the operation of machine learning models, model update is a fundamental process that requires careful consideration of its impact on downstream decision-making. Particularly when operating explainable models, changes in explanations resulting from model updates can lead to detrimental outcomes for users. Decision trees, due to their high transparency, are frequently employed in risk-sensitive decision-making and serve as a prominent example in which the aforementioned issue is evident. However, existing research addressing similar issues has focused on explanations based on feature contributions, and thus cannot handle explanations derived from tree structures. Therefore, this paper proposes the Backward Compatibility Loss in Tree-based eXplanations (BCLTX), a loss metric that suppresses changes in decision tree explanations before and after updates. Furthermore, we design CART with Backward Compatibility in Tree-based eXplanations (CART-BCTX), a lightweight algorithm that improves upon CART for the decision tree update problem under BCLTX. Experimental results using 10 real-world datasets, including both classification and regression tasks, show that CART-BCTX achieves favorable trade-offs between prediction performances and BCLTX values, with comparable computation times to CART, regardless of the task.

This work explores the relationship between task similarity and catastrophic forgetting in reinforcement learning. Catastrophic forgetting, the phenomenon in machine learning of losing the ability to effectively perform on previous tasks, is a significant impediment to continual learning. This study aims to understand the extent to which the similarity of a new task influences the performance on the previous task. Interpretable reinforcement learning, specifically Q-learning, is employed on graph-based tasks with the objective of minimising the number of steps to reach a goal. The study investigates the performance on a previously learned task after training on a new task, for tasks of varying relative levels of complexity. The experimental results reveal a complex dynamic between task similarity and forgetting, with significant fluctuations in forgetting severity observed across degrees of task similarities and task complexities, and are suggestive of an interdependence of forgetting on the similarity and complexity of tasks. The observations were accompanied by observations of high degrees of variability in forgetting and an uneven distribution of task similarity measures. The relationship between these variables remains unclear and no evidence of statistical significance that task similarity has an effect, independently, on forgetting is found in continual reinforcement learning. Further research is warranted to gain a comprehensive understanding of the potential interplay between task similarity and catastrophic forgetting.

The single-selection prophet inequality is a canonical Bayesian online selection problem in which independent nonnegative values arrive sequentially and the decision-maker must irrevocably select at most one. Classical single-threshold guarantees are tight in the worst case, but the hard instances that prove tightness are highly irregular: the prophet's advantage is driven by rare, very large realizations of the maximum. We refine this worst-case picture by imposing a bound on the relative variance of the prophet's value, $\mathrm{Var}(\max_{i\in[n]}X_i)/\mathbb E[\max_{i\in[n]}X_i]^2$. This yields a nonparametric complexity measure that interpolates between deterministic instances, where the full prophet value can be recovered, and the unrestricted worst-case regime. Our main technical contribution is a general kernel method for single-threshold prophet inequalities. The method represents an instance by the quantile function of the maximum and rewrites the payoff of a threshold as a linear kernel functional of this quantile. This turns the worst-case analysis into an infinite-dimensional convex program, restores strong minimax duality in quantile space, and reduces the bounded-variance adversary's problem to a one-parameter variational family. Applying this framework, we obtain an exact characterization of the IID bounded-variance curve and asymptotically optimal finite-horizon thresholds, a closed-form expression for the fixed-order non-identical model, and a prophet-secretary lower-bound program together with a strict separation from the IID benchmark at every positive finite variance constraint. As a further application of the same kernel viewpoint, we derive an exact formula for IID random horizons under a convexity condition on the horizon pgf, which includes monotone-hazard-rate horizons, highlighting the broad applicability of this new technique for single threshold settings.

How much an AI coding agent costs to run can depend more on the agent scaffolding that drives it than on the interface through which it reaches its tools. We set out to measure the cost of tool use over the Model Context Protocol (MCP) against tool use over an ordinary command-line interface (CLI), a difference on which published estimates disagree by more than an order of magnitude while resting on practitioner reports that cannot be reproduced. We ran one fixed software task -- six operations against a private online git repository -- across seven agent scaffoldings and five language models, and we verified completion by inspecting the repository state rather than trusting the agent's self-report. The dominant effect was the scaffolding. Two of the seven ship no MCP support at all; they completed every run using only the CLI, which shows that MCP is unnecessary for this class of work, and they were 5.0x to 28x cheaper than the five scaffoldings that do support MCP, comparing CLI runs alone with no MCP server attached anywhere. The effect was largest for a small 27-billion-parameter model running locally, whose cost varied 139x across scaffoldings while it completed the task under all of them. The comparison we set out to make proved unstable: thirteen strictly paired MCP-to-CLI ratios span 0.43x to 29x, with outliers on both sides. The two interfaces separate on the cost of failure, where 12.9 per cent of the money spent on MCP runs bought no completed work against 2.2 per cent on CLI runs, but not on its frequency: failures were equally common in both, in the original runs and in their repetitions alike. Agents frequently ignored the interface they were assigned, so comparisons that do not verify actual behaviour measure an unknown mixture. The harness, the task, the verification and the complete dataset are released as open source.

We present \LegoLM{}, a structured weight-sharing compression framework for large language models grounded in a systematic study of why global weight sharing fails and how to fix it. We identify two distinct failure modes. Distributional mismatch: for vector blocks of dimension d <= 2, transformer layers with heterogeneous weight scales impose a scale-mismatch penalty that grows linearly with d and cannot be resolved by increasing K, producing perplexity in the millions.Outlier dominance: for scalar blocks, a fraction ~1/K of weights lies beyond the outermost Lloyd-Max decision threshold and cannot be represented by any centroid; their misrepresentation accumulates across layers, causing catastrophic quality loss. \LegoLM{} resolves both failure modes via three data-free adaptations: 1 scalar-block encoding to eliminate the $d$-linear mismatch component, 2 percentile-selective replacement that identifies and preserves outlier weights verbatim, and 3 boundary-layer protection for the first and last transformer blocks. Across GPT-2 small (124M) and Mistral-7B, \LegoLM{} achieves +0.03% PPL degradation at 4.41X compression on Mistral-7B - outperforming PTQ-8bit in both quality and compression ratio - and -0.02% at 2.67X. Downstream evaluation on LAMBADA and HellaSwag confirms that \LegoLM{} at K=64, p=99% preserves accuracy within noise at 5.12 X compression, exceeding PTQ-8bit's compression ratio while matching its accuracy. We further discover that outlier dominance grows with model scale: full replacement at K=128 degrades GPT-2 small by only +23% but catastrophically degrades Mistral-7B by +1,134,279%, while selective replacement at p=99% rescues both models to under +15%. A controlled ablation confirms that selective replacement is the dominant mechanism: adding it to per-layer K-means also yields near-lossless quality, matching \LegoLM{} within 0.02%.

Mixture-of-Experts models increase parameter capacity while keeping the computation activated by each token bounded, but their architectural evolution cannot be explained by a chronological list of model releases alone. This technical survey synthesizes primary papers, official technical reports, and prior surveys to organize modern Mixture-of-Experts systems along five coupled dimensions: expert granularity, expert topology, routing freedom, the scope of load balancing, and execution structure. We describe eight architectural milestones as a dependency graph with six mainline developments and two orthogonal branches, rather than as eight successive generations. We then analyze individual systems through four control planes: Expert Topology, Routing, Balance, and Expert Parallelism. These planes specify which experts exist, which experts process each token, how aggregate load is controlled, and how selected computation is mapped onto physical devices. The framework connects algorithmic choices such as Top-k routing, shared experts, fine-grained experts, and dynamic expert composition with systems concerns including token dispatch, device placement, all-to-all communication, and communication-computation overlap. We conclude with equal-budget pretraining experiments, quality and systems metrics, and open research questions. The main trend is a shift from merely activating more sparse parameters toward decoupling semantic routing, computational budgets, and physical execution.

Trajectory-User Linking (TUL) aims to identify the owner of an anonymous trajectory from a set of candidate users, providing a basis for user mobility analysis and personalized location-aware services. Existing methods often learn Point of Interest (POI), temporal, and semantic features independently, make limited use of structural knowledge shared across trajectories, and compress structural and sequential information before classification. To address these issues, we propose Multi-Relational Knowledge Graph Enhanced Embedding for Trajectory-User Linking (MakeTUL), which, to the best of our knowledge, is the first attempt to introduce knowledge graph representation learning into TUL. MakeTUL organizes visit-time, POI-category, and transfer-speed information as typed relations in a multi-relational mobility knowledge graph, allowing heterogeneous mobility semantics to jointly constrain the learned embeddings. The resulting POI representations are further enriched with high-order co-occurrence patterns extracted from the trajectory collection, providing structural prior knowledge for sparse and overlapping trajectories. By integrating these prior-enhanced representations with temporal, category, and transfer information, the trajectory sequence learning module captures ordered mobility patterns, while a dual-branch classification layer preserves and combines global structural evidence and sequential evidence at the decision level.

We consider the problem of sampling compositional and discrete objects from a given unnormalized posterior distribution. Notably, recent studies have shown that this problem can be efficiently solved by learning a deterministic Markov Decision Process (MDP) that progressively builds each object in proportion to the posterior. In this work, however, we demonstrate that the Markovian assumption can both hamper signal propagation during training and catastrophically reduce the learned sampler's expressivity due to state aliasing. To address these issues, we propose lifting the MDP with a learnable latent dynamical system that allows the underlying policy to depend on the entire past trajectory---and not only on the current state. In view of this, we refer to the resulting method as path-dependent discrete amortized inference. Importantly, we provably extend existing learning algorithms for discrete amortized samplers to our setting. In experiments on standard benchmark problems, we also show that our approach often leads to faster learning convergence and improved state space exploration relatively to prior techniques.

We study exact Kullback--Leibler (KL) projection for low-rank factorizations whose two nonnegative factors have prescribed row marginals and a shared, learned column marginal. For arbitrary positive row marginals of equal total mass, the joint KL projection reduces exactly to a strictly convex gauge-fixed dual with only $r-1$ effective variables; its Hessian is a sum of categorical covariance terms and admits $O((n+m)r)$ matrix-free Hessian--vector products. The projection theorem is objective-independent. We then specialize this geometry to doubly stochastic (DS) graph learning through $W=U\operatorname{Diag}(g)^{-1}V^\top$, where row-simplex factors with a common column mass induce an exactly DS graph without materializing an $n\times n$ optimization variable. Combined with observed-edge sparse fitting, a stochastic anchor-reduced manifold regularizer, and Bregman backtracking, the resulting mirror-descent method preserves exact feasibility at every accepted step. Under a nonvanishing latent-mass condition, it satisfies sufficient decrease and an $O(1/N)$ mirror-stationarity bound, while strictly positive accumulation points are KKT stationary. Matched clustering experiments show competitive accuracy, feasibility residuals near numerical precision, and favorable anytime behavior without a dense learned graph.

Large language model agents increasingly rely on reusable skills to extend their capabilities beyond parametric knowl- edge. However, retrieving the appropriate skill from a large- scale library remains challenging because realistic user re- quests are often concise and underspecified, stating only the task goal while leaving the required capabilities and execu- tion steps implicit. Existing benchmarks provide limited cov- erage of such requests. To address this gap, we introduce SkillReason-Bench, a large-scale cross-domain benchmark containing 3,729 queries and a retrieval corpus of 61,228 skills spanning nine domains. We further propose SkillRea- son, a two-stage framework that uses chain-of-thought rea- soning as training-time supervision for skill retrieval. In Stage I, capability reasoning traces generated by a stronger teacher provide explicit supervision through contrastive learning, re- trieval distribution alignment, and language modeling, en- couraging the retriever to internalize capability reasoning in its query representation. In Stage II, a retrieval-guided GRPO objective encourages the model to explore reasoning trajecto- ries better suited to its own capabilities and more effective for retrieval. At inference, SkillReason directly encodes the orig- inal query without autoregressive CoT generation, preserv- ing efficient query-only retrieval. Extensive experiments on SkillReason-Bench, SkillRet, and SRA-Bench show that Skill- Reason achieves state-of-the-art performance across all three benchmarks, demonstrating that reasoning-enhanced training better bridges the semantic gap between high-level task goals and skill capabilities.

Open lakehouse table formats accumulate small data files over time, which degrades query performance. Deciding when compaction is worthwhile remains threshold-driven, but which metadata features actually determine compaction utility is not well understood. We present an open simulation framework that generates 2,376 Apache Iceberg tables spanning three orders of magnitude in file size, extracts 17 metadata features from manifest files without reading data, and trains XGBoost to predict the continuous file-reduction ratio (R2 = 0.998, RMSE= 0.013). The binary compaction decision turns out to be trivially separable by a single partition-level threshold max_files_per_partition> 4, requiring no learned model. Cross-schema validation on 96 TPC-H tables confirms generalisation without retraining (R2 = 0.976). A query benchmark reveals that compaction benefits metadata-heavy queries but can slow full-scan aggregations by reducing task parallelism. All code and data are publicly available.

Zero-shot text-to-speech (TTS) now supports interactive assistants, personalized media, and accessibility tools. All TTS systems require faithful linguistic rendering, consistent speaker identity, and low-latency response. Yet compact streaming systems must preserve sufficient acoustic detail in a predictable low-rate latent sequence, while iterative diffusion sampling and classifier-free guidance multiply inference cost at every autoregressive step. To strike a balance between high-fidelity synthesis and low-latency inference, we present CuteTTS, a compact continuous-autoregressive TTS system. It combines semantically aligned causal VAE latents with patch-level autoregression, explicit speaker conditioning, and a bidirectional flow-matching head. We further introduce guidance-step distillation, which absorbs classifier-free guidance and multiple solver steps into a single interval-conditioned student. Evaluations on LibriSpeech and Seed-TTS-Eval demonstrate competitive intelligibility and speaker similarity in zero-shot voice cloning, while distillation lowers first-audio latency by 23.3% and real-time factor by 40.8% relative to the base model with comparable objective and subjective quality. These results provide a practical path toward continuous-autoregressive TTS that reconciles high-fidelity generation with the latency demands of real-time interaction.

Fuzzing Rust library APIs requires constructing well-typed, compilable call sequences that satisfy ownership rules, generic parameters, and trait bounds; existing tools ignore these constraints or use shallow heuristics, yielding low coverage. We present GRAFT, which extracts structured API information from Rust documentation, builds an API dependency graph via recursive generics-aware type matching, and uses topology-guided traversal plus LLM synthesis with compiler-error feedback to produce compilable fuzz targets. On 13 crates from crates.io, GRAFT achieves 80.75% macro-average API coverage at 96.19% compilation success, outperforming RULF and RPG by 4.76x and 2.43x, and reaching 1.41x the average API coverage of deepSURF on crates with unsafe-reaching APIs.

Scientific named entity recognition (SciNER) plays a crucial role in information extraction and knowledge discovery from scientific texts. Recently, large language models (LLMs) have demonstrated the capacity to achieve competitive SciNER performance with minimal human effort. Existing research highlights the importance of incorporating candidate entity type information for accurate entity recognition and classification by LLMs. However, when too many candidate entity types are provided in the prompt, LLMs struggle to accurately recognize and label entities in scientific texts, where entity types are more complex than in general domains. To address this challenge, we propose TdSciNER, a type-driven approach that effectively leverages entity type information to enhance SciNER performance. In TdSciNER, we first design an entity type filter model to identify the most likely entity types present in a given sentence. Subsequently, we introduce an auxiliary multi-class entity typing task within a multi-task learning framework alongside SciNER to obtain richer contextual representations. Then, we develop a novel demonstration selection strategy based on sentence similarity and entity type diversity to activate the in-context learning capabilities of LLMs, thereby improving entity recognition accuracy across diverse scientific domains. Experiments on three datasets demonstrate that our method achieves performance comparable to fully supervised models. Further analysis validates that each entity type-driven component in TdSciNER contributes to the improvement of SciNER performance. This work provides valuable insights for future advancements in SciNER and broader information extraction tasks in scientific text mining.

Open-weight language models from Chinese AI labs caught up on benchmarks relative to proprietary frontier models in recent months. Yet their reliability on real-world financial tasks remains largely untested. We updated the Financial Touchstone benchmark, which now has 2,967 question context-answer triplets across 495 international annual reports. We also apply a new set of models on the benchmark, expanding coverage from eleven to twenty models across ten providers, including recent open-weight models such as GLM 4.7, GLM 5, Kimi K2.6, and DeepSeek V3.2, as well as Alibaba's proprietary flagship Qwen3-Max. Anthropic's Claude Opus 4.6 achieves the highest accuracy (88.4%), while Google's Gemini 2.5 Pro maintains the lowest hallucination rate (0.08%). Notably, the open-weight Kimi K2.6 ranks third in accuracy, and the non-reasoning models GLM 5 and Mistral 3 rank fourth and fifth, challenging the assumption that reasoning architectures or proprietary weights are a prerequisite for strong financial comprehension. Information retrieval remains the primary bottleneck, accounting for 48.9% of all failures. We also document a new finding: geopolitical content filters in Chinese models refuse legitimate financial questions (0.08% of attempts), sometimes without clear reason, and the refusal behavior depends on the access route as much as on the model. The complete dataset and evaluation framework are publicly available.

Airport landside traffic centers connect terminal arrivals with taxis, ride-hailing vehicles, private cars, buses, metro services, parking facilities, and terminal-area roadways. Peak arrivals can create coupled congestion across passenger queues, vehicle queues, pickup berths, storage areas, and access roads. This study proposes a QUBO-inspired computational framework for bottleneck diagnosis and dynamic dispatch in this setting. Shanghai Pudong International Airport and Hangzhou Xiaoshan International Airport serve as case airports. A five-minute state model links passenger arrivals, vehicle supply, pickup berth service, vehicle storage, and road capacity. Bottleneck diagnosis uses service intensity, road demand saturation, bottleneck frequency, queue severity, shadow-price leverage, and a composite congestion severity index. Two dispatch schemes are tested under consistent demand inputs: finite-action model predictive control and quadratic-unconstrained-binary-optimization-inspired simulated annealing. In the strong-peak baseline scenario, the QUBO-inspired method reduces the final passenger queue from 3445 to 2477 passengers at Shanghai Pudong and from 2053 to 1482 passengers at Hangzhou Xiaoshan. Case results indicate different dominant bottlenecks. Shanghai Pudong is more affected by road saturation, whereas Hangzhou Xiaoshan is more affected by pickup berth service. Robustness tests under demand, supply, service, road-capacity, modal-share, and random-noise perturbations show retained queue-reduction benefits under the tested uncertainty levels.

Digital-twin calibration requires interaction data that is expensive to collect. We study two acquisition decisions: which trajectories to generate, and when to spend a limited budget on privileged parameter measurements. Our framework couples an excitation-oriented reinforcement learning controller, a recurrent parameter estimator with predictive uncertainty, and a budgeted query policy. In Pendulum, a Random Forest diagnostic recovers gravity only weakly from task-oriented trajectories and does not recover mass or length, while a GRU trained on excitation-oriented trajectories reaches a mean absolute error of 0.0066 with no queries. We then withdraw continuous oracle access partway through an episode, so that the twin must run on the estimator's output for the remainder. The estimator-plus-policy pipeline achieves a terminal error of 0.0092 under a three-query budget, against 0.2031 for an uncalibrated twin. In partially observable Waterworld, five controllers produce different observed error profiles across three hidden parameters, and an estimator trained on a five-controller mixture reaches online normalized errors of roughly 4-5%. These exploratory case studies are not controlled ablations, but they motivate treating trajectory design and query allocation as explicit design variables in data-scarce calibration.

Sparse mixture-of-experts (MoE) layers expand recommendation capacity through conditional computation, yet a trained checkpoint still stores and routes over its full expert bank. We study a deployment problem: convert that checkpoint to a smaller standard MoE under an explicit expert budget, without adding a compression-specific online module. To address this, we introduce UniMoMo, a post-training compression framework formulated as a constrained graph coarsening problem. Rather than relying on parameter distance, UniMoMo groups experts based on their functional similarity, using an unlabeled calibration set to measure how similarly experts respond to shared recommendation states. To prevent performance degradation, we introduce a layer-adaptive protection mechanism that restricts the merging of high-traffic experts based on their routing exposure. Across Amazon Beauty, KuaiRec, and TenRec with 2, 4, and 6 MoE blocks, the final four-expert checkpoints obtain source-relative five-run mean NDCG@10 ratios of 99.92%--102.30% and measured A100 speedups of 1.28$\times$--1.63$\times$. An aggressive two-expert, top-1 operating point obtains ratios of 98.36%--104.24% and speedups of 1.47$\times$--2.21$\times$. These endpoint results evaluate the complete conversion-and-adaptation workflow and show that a trained recommendation MoE can be exported at multiple serving budgets.

Domain generalization (DG) and neural network pruning are conventionally treated as distinct objectives, targeting out-of-distribution (OOD) robustness and model efficiency, respectively. In this work, we bridge this gap by introducing Domain-Aware Pruning (DAP), a framework that leverages network sparsity as a mechanism to implicitly enhance generalization to unseen domains. Diverging from standard binary mask optimization, DAP learns a continuous parameter retention probability $p \in [0, 1]$, framing network compression as a continuous probabilistic masking problem. By introducing a regularization objective that actively penalizes the retention of domain-sensitive weights during the mask training, DAP identifies a domain-invariant subnetwork. Empirical results across five DG benchmark datasets demonstrate that DAP achieves significant sparsity while consistently matching or exceeding the OOD performance of its dense counterparts. Crucially, DAP is an algorithm-agnostic framework that integrates seamlessly with existing DG pipelines without necessitating post-hoc fine-tuning. Beyond efficiency and generalization, we show that DAP natively provides increased robustness to adversarial perturbations and yields highly interpretable models, where the retained weights reliably encapsulate the most domain-invariant and task-critical representations.

In Reinforcement Learning with Verifiable Rewards (RLVR) frameworks for mathematical reasoning tasks, floating-point results are typically evaluated using a tolerance-based reward. However, this strategy suffers from challenges such as difficulty in threshold calibration, unstable training dynamics, and limited accuracy, especially in clinical scenarios. To address these limitations, we propose a knowledge-guided hybrid reward framework (\textsc{MedCalc-R1}). Specifically, we introduce a knowledge verification reward mechanism that enforces explicit generation of computational formulas, which are further validated by an external verifier to enhance interpretability and reasoning reliability. Furthermore, we design a hybrid soft-hard reward scheme combining a hard constraint based on clinical safety thresholds with a soft, precision-sensitive reward that progressively guides learning within the acceptable range. Experimental results demonstrate that our method significantly outperforms existing baselines in both reasoning accuracy and generalization capability, validating the effectiveness and applicability in safety-critical domains.

Running a business is a challenging form of intelligent work. Operators must infer opportunities from partial signals, commit capital under uncertainty, adapt to delayed outcomes in a changing market, and satisfy regulatory obligations before trading legally. Frontier LLM agents can increasingly complete complex workflows, yet business-related capabilities are rarely evaluated in existing agent benchmarks. We introduce \textbf{Business Arena}, a controlled environment where an AI agent runs a cross-border shop, buying from suppliers and selling to buyers over a long horizon. We ground the arena in real Alibaba.com sourcing data and market conditions calibrated from authoritative sources. Delayed and coupled consequences make individual business decisions difficult to judge, but their combined outcome is measurable through profit. Because profit alone cannot explain why an agent succeeds or fails, we compare agents with human-designed strategies to estimate available opportunity, use skill-level metrics to reveal underlying strengths and weaknesses, and trace realized gains and losses to the actions that produced them. We use mechanism ablations to establish that strong results reflect genuine business intelligence rather than neglect or simulator-specific shortcuts. We evaluate 15 frontier models and find a ninefold difference in mean final net worth. Even the best model falls behind human-designed strategies, indicating that business operation remains challenging for LLM agents. Skill-level analysis reveals operating styles, from margin-focused premium sellers to high-turnover wholesalers and customer-service specialists, while action-level attribution identifies the sourcing, pricing, and recovery decisions that create or destroy value. Together, Business Arena takes a first step toward a realistic and trustworthy testbed for evaluating end-to-end business agents.

Industrial device commissioning requires engineers to manually extract hundreds of protocol-specific parameters from heterogeneous PDF manuals and transcribe them into supervisory control systems, a time-intensive, error-prone workflow. This paper presents SysName, a production-oriented pipeline that automates device configuration end-to-end for Modbus RTU, OPC-UA, Profibus DP, and CANopen. It builds a hybrid dense-sparse retrieval index augmented by an ontology graph derived from ECLASS, AAS, and SOSA/SSN, using a BGE-M3 encoder with a cross-encoder reranker to surface relevant manual passages. A local LLM (T=0.1) generates ontology-aligned JSON-LD configurations via protocol-specific prompts and a four-step repair pipeline. A two-stage abstention gate, combining a reranker-score threshold and an IRI resolution ratio, blocks unsafe LLM invocations and filters low-coverage configurations before SHACL validation. On a gold set of 28 field-level queries, the hybrid retriever reaches 0.96 HitRate@10, and the reranker raises MRR@10 from 0.56 to 0.63 with perfect score separation for abstention. The generator attains field-level F1=0.87 with exact match on 9 of 12 runs. End-to-end runs on an H100 GPU complete in 2.6-6.6s per device with zero unsafe writes and zero silent failures on a five-device benchmark; every unsuccessful run is flagged by abstention or deployment verification. Component-wise evaluation localises the single systematic failure to OPC-UA generation, invisible to end-to-end metrics alone. A case study commissions a physics-simulated Universal Robots UR5e robot from unmodified vendor documentation (254-page manual, 8-page register list, 496 chunks), reaching field-level F1=1.0 over three runs with read-back and joint-consistency verification. An ablation study and comparison with five industrial-LLM systems complete the analysis.

Group discussion-based teaching is widely used to foster collaborative learning, yet teachers in physical classrooms often struggle to simultaneously monitor multiple groups and quickly diagnose a target group before intervening. Existing visual analytics tools primarily support post-hoc analysis on desktop, providing limited support for in-situ walk-around teaching. To address this gap, we present MobileGroupVis, a mobile visual analytics system for in-situ analysis of classroom group discussions. MobileGroupVis integrates multi-group monitoring, single-group diagnosis, and instructional intervention into a concise analytical workflow tailored for small-screen touch interaction. The system is powered by a lightweight streaming analysis pipeline that converts group audio into structured discussion data and further extracts interaction patterns, topic progression, and topic deviation through a dialogue analysis module. To enable both glanceable overview and traceable diagnosis, we design six coordinated views, including a compact glyph that visually encodes word count, interaction intensity, and topic deviation for efficient cross-group comparison and anomaly localization, along with detailed views for opinion evolution, interaction dynamics, topic coverage, and dialogue records. We evaluate MobileGroupVis through two case studies and expert interviews. The results provide preliminary evidence that MobileGroupVis supports teachers in understanding discussion processes, identifying groups in need of attention, and facilitating in-class intervention.

Recently, retrieval-augmented and memory-augmented methods have emerged as two promising paradigms for long-video question answering. However, existing methods typically rely on rigid, fixed-length temporal chunking (e.g., 10s) and static offline memory banks, which not only fragment coherent continuous events but also fail to adapt during real-time reasoning. Moreover, whether using multi-scale summaries or multimodal knowledge graphs, current approaches prioritize retrieval relevance while overlooking evidence sufficiency, often stopping to answer once only semantically relevant clues are retrieved, even when key temporal, causal, or fine-grained action evidence is still missing. To tackle these challenges, we propose REVEAL, a rubric-guided agent framework. As a foundation, we introduce an adaptive visual-similarity-based preprocessing pipeline that groups visually coherent adjacent frames into natural event units to construct an offline-online video memory---capturing global video context offline while dynamically maintaining question-conditioned memory online. Built upon this structured memory, REVEAL uses an automatically constructed rubric library to explicitly verify whether retrieved evidence satisfies sufficiency criteria, pinpoints missing clues upon verification failure, and directs targeted re-retrieval for complementary information. Without any extra training, REVEAL consistently outperforms both closed-source and open-source state-of-the-art methods across extensive experiments. These results show that explicitly verifying evidence sufficiency, rather than stopping at semantic relevance, retrieves the decisive clues that prior methods miss and yields more reliable long-video reasoning.

Fourier neural operators (FNOs) provide efficient nonlocal spectral learning, but varying geometries and independently chosen discretizations remain difficult to accommodate. We introduce the ambient-domain extension Fourier neural operator (ADEx-FNO), a deterministic framework that incorporates geometry without modifying the defining Fourier-operator layers. Each physical domain is embedded in a fixed ambient hypercube and represented by a signed distance function. Inputs and solution fields are deterministically extended to the ambient domain, transferred to a common, potentially nonuniform rectilinear latent grid, processed by the FNO, then interpolated to an independently chosen target discretization and restricted to the physical domain. All geometry-transfer operations lie outside the optimization procedure and require no trainable graph, point-cloud, deformation, or geometry-decoding modules. ADEx-FNO achieves relative l2 errors of 0.32%-0.77% on held-out smooth-domain nonlinear Poisson and advection-reaction-diffusion problems in 2D and 3D, and is also evaluated on unseen nonsmooth geometries. A single ADEx-FNO inference is then used to initialize conventional CFD solvers. For all 29 converged 2D and 3D RANS cases, pseudo-time iterations decrease, with mean reductions of 44.17% and 43.03%, respectively, with comparable gains across three mesh resolutions. URANS cases reduce post-window physical-time advances by 18.52%-27.51%. In transfer from 2D URANS training data to DNS at different Mach and Reynolds numbers, the bootstrap interval decreases by 23.47%-48.21%, depending on the target statistic. In all CFD tests, ADEx-FNO provides only the initial field; the governing-equation solver controls the subsequent solution, while physical or statistical consistency is assessed separately from computational savings.

Mental health disorders are a leading cause of disability worldwide, yet Natural Language Processing (NLP) research for mental healthcare has remained concentrated in high-income, English-language settings. North Africa, and Algeria in particular, is largely absent from this literature despite its unique linguistic, historical, and healthcare context. We present the first conceptual framework examining the potential role of NLP within Algeria's mental healthcare system. Drawing on narrative synthesis of global NLP mental health research, Algerian healthcare literature, and low-resource NLP methodologies, we identify four structural barriers to mental healthcare: the language-of-care gap, geographic inequities in access, stigma-related barriers to help-seeking, and the absence of research and digital infrastructure. We then map existing NLP capabilities to each barrier, outlining their potential applications, implementation constraints, and the technical, institutional, and governance requirements necessary for deployment. Based on this analysis, we propose a research and policy roadmap that prioritizes data resources, multilingual language technologies, evaluation frameworks, and regulatory capacity. Although grounded in the Algerian context, the framework addresses challenges common to many multilingual, low-resource, and post-colonial settings. This work provides a foundation for future research on culturally and linguistically appropriate NLP for mental healthcare and offers a practical roadmap for developing responsible AI-enabled mental health systems in underrepresented regions.

Machine translation (MT) systems often fail to correctly translate gender, especially when converting from a gender-neutral language like English to a gendered target language such as Romanian. This bias results in translations that default to masculine forms or reinforce gender stereotypes. We propose a hybrid pipeline to mitigate this issue by combining large language model (LLM)-based gender classification with neural machine translation (NMT). Our system uses a fine-tuned LLM to detect the intended gender of target words in English sentences and insert inline gender hint tags. These tagged sentences are then passed to a Transformer model fine-tuned to generate morphologically correct Romanian translations. To support this, we introduce three novel datasets for gender disambiguation and translation. Our approach improves gender accuracy on the WinoMT and WinoGender benchmarks by over 40 percentage points compared to a baseline MT system. This is the first method to explicitly address and evaluate gender bias in English-Romanian MT using both LLM inference and tag-aware translation.

Multi-agent systems (MAS) built on Large Language Models (LLMs) are proliferating rapidly, but their heterogeneous execution traces provide no common basis for evaluation across methods. Outcome-only benchmarks discard collaborations, whereas LLM-as-Judge evaluation requires additional, model-dependent inference and can vary with the LLM and rubric. We introduce a generalizable evaluation framework that maps native MAS traces into a shared space of unified collaboration graphs, enabling different methods to be evaluated under the same representation, reference set, and metric panel. Candidate graphs are compared with a query-specific reference forest. Each forest is a benchmark-provided collection of verified-success graphs: it records diverse ways in which representative MAS methods can complete the task, rather than prescribing a unique optimal process. Instantiating the framework as ForestBench, we filter $844$ collaboration-necessary queries from seven public datasets, precompute ten successful target-conditioned reference graphs per query, and evaluate six representative MAS frameworks. Controlled backbone, reference-construction, and perturbation studies test the stability and scope of evaluation. Once the benchmark forests are built, ForestBench scores a trace in milliseconds without further LLM inference, providing a reusable structural basis for comparing diverse MAS collaboration traces.

Multi-agent reinforcement learning (MARL) is a powerful framework for solving complex collaborative tasks, but it relies heavily on well-defined global reward functions. Designing such rewards is challenging, especially in systems with heterogeneous agents, where a single scalar objective may fail to capture diverse behaviors. In this paper, we introduce Multi-AGent Preference-Integrated lEarning (MAGPIE), which addresses these challenges through agent-specific preference modeling. Each agent is evaluated by a dedicated expert through preference signals, eliminating the need for global evaluation. We theoretically prove that optimizing these decentralized preferences converges to a Nash equilibrium policy. To integrate local preferences into a coherent global objective, we construct agent-specific reward models from preference data and combine them via a monotonic aggregation mechanism. We further prove that optimizing this aggregate reward model is equivalent to training the Nash equilibrium policy. Extensive experiments on benchmark multi-agent tasks and a sequential production line task show that MAGPIE achieves performance comparable to reward-engineered baselines, demonstrating its potential to facilitate policy learning in scenarios where precise reward engineering is impractical.

To anticipate socio-technical risks from AI agents, organizations need taxonomies to classify them. However, existing AI risk taxonomies focus on broad risks and do not capture job-specific risks introduced by agents. To address this gap, we make three main contributions. First, we developed a multi-layer framework from a literature review of AI agents. The framework models three core components and their interactions: agents, goals, and environment. Second, we embedded this framework in a structured prompt and applied it to descriptions of 2,078 job tasks from the O*NET database, producing 8,356 risk scenarios labeled by severity and deployment mode (automation or augmentation). We validated these scenarios with 45 workers across 10 job roles and an independent LLM judge, confirming their plausibility and alignment with job tasks. Finally, we extended an existing taxonomy to create a 15-category taxonomy of workplace AI agent risks that covers all our risk scenarios. Our analysis highlights four findings. First, augmentation is not inherently safe because overreliance on agents can gradually erode workers' skills and oversight. Second, Erroneous Agent Actions accounts for the largest share of risk scenarios and has the highest concentration of severe risks. Many arise at the human-agent boundary. Third, automation is associated mainly with organizational risks, while augmentation is associated mainly with risks to workers. Fourth, workers found our taxonomy easier to use for a risk classification task than two other taxonomies and preferred it in 64% of non-tied comparisons with a recent generative AI risk taxonomy. These findings show that workplace AI agent risks do not arise from agents alone; they also depend on how people work with agents and how agents are deployed. Safer workplaces require not only safer agents but also carefully designed human-AI agent collaboration.

World models have recently achieved impressive progress in visual prediction and interactive generation, but extending them to multi-agent environments introduces a fundamental scalability challenge. Existing methods generally assume a fixed number of agents during training and inference, which ties the model to a pre-determined agent population and limits inference-time scalability. Our key insight is that cross-view consistency should arise from a shared world state whose evolution does not assume a predefined number of agents, while agent-specific observations should be generated by querying this state through a unified rendering interface. Based on this insight, we propose Khora, a scalable multi-agent world model that supports inference-time expansion to arbitrary numbers of agents without retraining. Our framework decouples world-state evolution from visual rendering and introduces a population-agnostic rendering mechanism for incorporating other agent information. This design maintains cross-view consistency through the shared world state rather than through dense interactions among observation streams inside the expensive video generator, enabling approximately linear practical scaling with the number of queried views. Qualitative experiments demonstrate that our approach generalizes to unseen numbers of agents while maintaining visual quality and multi-agent consistency. We further implement a real-time interactive system to demonstrate scalable open-world simulation.

Diffusion bridge models leverage Doob's \(h\)-transform to construct stochastic transports between arbitrary endpoint distributions, and have shown strong potential in image-to-image translation and restoration. However, most existing bridge models rely on hard endpoint conditioning, which forces the terminal state to match a prescribed target exactly. This hard constraint induces terminal-boundary singularities: the terminal law collapses to a Dirac measure, and the resulting drift coefficients become ill-conditioned near the endpoint. In this paper, we propose Soft Denoising Diffusion Bridge Models (SDDBMs), a generalized framework that regularizes diffusion bridges directly at the level of their terminal constraints. Instead of imposing an exact endpoint, SDDBMs prescribe a non-degenerate Gaussian terminal marginal under the transformed path measure, with a flexible terminal center and variance. Starting from this prescribed marginal, we develop a complete closed-form construction of the soft bridge, including the Gaussian terminal reweighting and soft \(h\)-function, the induced Gaussian forward marginals and \(\mathbf{x}_0\)-free dynamics. Theoretically, SDDBMs provide a unified probabilistic perspective that encompasses existing diffusion bridge models, including DDBMs, GOUB, and UniDB, as special cases under specific parameter choices. Extensive experiments on image restoration tasks demonstrate that SDDBMs achieve improved numerical stability and superior generation quality over existing bridge-based methods.

Hamiltonian Monte Carlo (HMC) is a successful generic inference method in probabilistic programming, but in its ordinary formulation it needs gradients and finite-dimensional parameter spaces. In Haskell, lazy evaluation lets probabilistic programs express stochastic processes and other non-parametric Bayesian models over implicit infinite-dimensional spaces. This paper develops new formulations of gradient-based HMC for this infinite-dimensional setting, via lazy evaluation. For automatic differentiation, we provide an analysis based on a new notion of "piecewise analytic under cylindrical analytic partition" (PACAP), to show that even if a program is infinite-dimensional and defined lazily, the gradient of the likelihood function is finitely supported. For the Monte Carlo method itself, we develop several HMC variants and a No-U-Turn Sampler that operate over the infinite-dimensional parameter space but are still productive because of lazy evaluation. Experiments cover Gaussian mixture clustering, random walks, and piecewise-constant regression with Poisson-process changepoints.

Fraud operations must allocate events among automatic approval, analyst review, and automatic blocking even though the labels needed to evaluate these actions are selective and delayed. Predictive scores order cases, but they do not show whether the evidence is current and representative enough to delegate an action to the model. We develop freshness-constrained audit capacity (FCAC), a decision-support framework that treats automation as an authorization decision constrained by action risk, evidence freshness, and shared review capacity. It evaluates candidate action regions from mature randomized audits and a prespecified temporal allowance. Supported regions are automated; unsupported regions remain in review. The resulting decision record reports evidence age, audit demand, total review workload, value exposure, and compatible temporal change. We show that current action risk is unidentified without restricting unobserved label evolution. Under representative randomized audits, label-independent evidence windows, and a prespecified condition linking historical and current action risk, we derive simultaneous finite-sample control of unsafe authorization. Chronological evaluations with simulated audits on IEEE-CIS, ULB-Worldline, and Elliptic++ yield zero-drift automation rates of 84.4%, 67.4%, and 81.3%, with total review workloads of 24.1%, 46.0%, and 43.1%. The experiments reveal an audit-capacity trade-off: sparse auditing delays authorization, whereas intensive auditing eventually increases workload. A separately specified BAF stress test further indicates that fallback thresholds must reflect candidate-specific evidence rather than a common fraction of the risk limit. These findings identify audit freshness and analyst capacity as joint design considerations for fraud decision support.

Accurate 3D medical image segmentation requires the integration of long-range anatomical context with fine boundary detail. Existing methods often model global and local features in separate modules or feature levels and perform channel recalibration independently. This may cause semantic mismatch between global context and local boundaries, insufficient channel relationship modeling, weak spatial-channel interaction, and redundant representations. We propose CDGC-Net, a 3D medical image segmentation network that combines cooperative dual-scale spatial attention with grouped hierarchical channel modeling. With-in each CDGC block, Cooperative Dual-Scale Self-Attention (CDSA) assigns attention heads to parallel local-window and global-sparse branches. The two branches capture fine spatial details and long-range anatomical context at the same feature level. Their outputs are concatenated into an $N\times C$ spatial representation and directly passed to Grouped Hierarchical Channel Attention (GHCA). GHCA organizes the channels into $r$ groups and models both within-group and cross-group dependencies. CDSA and GHCA reuse a shared key projection to maintain a consistent feature reference. Residual feature alignment subsequently integrates the refined features with the original representation. On the Synapse, ACDC, BraTS, and LA datasets, CDGC-Net achieved mean DSC values of 86.96\%, 92.91\%, 82.56\%, and 93.52\%, respectively, exceeding the next-highest reported values by 0.39, 0.47, 0.17, and 0.32 percentage points. CDGC-Net contains 25.83M parameters and 28.62G FLOPs for an input size of $64\times128\times128$, reducing these quantities by 39.87\% and 40.30\%, respectively, relative to UNETR++. These results indicate a favorable trade-off between segmentation accuracy and computational complexity.

Crowdsourced Federated Learning (CrowdFL) extends traditional federated learning by enabling open and heterogeneous participation through a crowdsourcing paradigm. In this setting, reputation-driven incentive mechanisms are commonly employed to guide worker selection and enhance trustworthiness. While such approaches improve participant reliability, existing frameworks largely overlook the quantification of their robustness against stealthy adversaries, particularly those capable of evading standard detection mechanisms. To fill this gap, this paper proposes R2CFL, a robust reputation-driven CrowdFL framework. R2CFL introduces a robust reputation model coupled with a nearest neighbor mixing (R2-NNM) defense mechanism that links reputation evolution with the filtering of updates during aggregation. The proposed mechanism prevents stealthy attackers from gradually accumulating trust and influencing future tasks. Experimental results demonstrate that R2-NNM matches or surpasses state-of-the-art Byzantine-robust and backdoor defense mechanisms against adaptive attackers. Furthermore, when integrated with existing detect-and-filter defenses, the proposed reputation model faithfully captures the statistical robustness of the underlying defense by producing reputation scores that closely reflect its true positive and false positive characteristics.

Rejection sampling fine-tuning (RFT) is widely used to train code agents by generating trajectories on verifiable software engineering tasks, retaining those that pass the tests, and fine-tuning on the successful rollouts. However, even strong code agents repeatedly fail on a substantial fraction of such tasks, and standard RFT simply discards these failures. The discarded samples are precisely the hardest and most informative ones, drawn from verifiable instances that are costly to curate. Stronger base models may reduce the number of failures, but the remaining hard cases still define the frontier for further improvement. We propose FailForge, an agentic framework that converts failed rollouts into training signal. For each failed instance, an agent diagnoses the failure from error feedback and execution traces, distills the diagnosis into a concise and actionable skill, and injects the skill into the agent context for a guided second attempt. Trajectories that succeed under skill guidance are folded back into the RFT corpus. Crucially, the skill is removed at training time, so the model internalizes the recovered behavior rather than relying on external hints at inference. FailForge recovers over 26% of previously failed instances at marginal additional cost, and training Qwen3.5-4B on the augmented corpus improves the SWE-bench Verified resolve rate by 6.6 points over a strong RFT baseline, with gains concentrated on the hardest problems.

Recent advancements in Speech Large Language Models have demonstrated remarkable capabilities in understanding complex audio tasks. Despite this progress, their long-context inference remains severely bottlenecked by prohibitive KV cache memory demands. Existing text-centric compression methods struggle here, often disrupting speech continuity or discarding crucial semantic cues. To address this, we propose VoxZip, a train-free, two-stage semantic-anchored KV cache compression framework. The first stage uses automatic speech recognition (ASR) transcriptions as explicit semantic anchors to temporally align, compress, and fuse audio tokens, significantly reducing the initial KV cache while elevating token information density. To further improve the compression ratio, the second stage employs a dynamic filtering strategy based on temporally decayed accumulated attention to evict non-essential tokens while mitigating early-token bias. Comprehensive evaluations on Qwen3-Omni across six diverse audio benchmarks demonstrate the superiority of our approach. VoxZip excels in long-audio reasoning and consistently maintains high-fidelity perception on short-form tasks. Notably, it sustains over 90\% of the uncompressed baseline performance even under an aggressive 20x KV cache compression in long-context scenarios. Furthermore, at a 4x compression ratio, VoxZip yields a 1.9x increase in inference throughput alongside a 3.3x reduction in peak memory overhead. Code and models will be available at https://github.com/MM-Speech/VoxZip.

A central obstacle in building graph foundation models is the input heterogeneity in terms of feature space dimensionality, semantics, and structure. Such heterogeneity limits the capability of graph neural networks to generalize to new graphs with unseen feature spaces. We address the transferability challenge with SIGIL, a framework that maps any attributed graph to a unified representation space of fixed dimension. Given a graph, SIGIL lifts it to a structural interaction graph, where nodes are the input feature dimensions and weighted, typed edges encode feature alignment across multiple orders of the graph's connectivity. A relational message-passing network embeds each feature dimension into a shared space, transforming the original node features, of arbitrary dimensionality, into representations transferable to any downstream graph. By construction, SIGIL is equivariant to permutations of nodes, feature dimensions, and labels. Additionally, when the input features are one-hot indicators of discrete relations, SIGIL recovers and strictly generalizes existing foundation models for knowledge graph reasoning. A single SIGIL model, pretrained on one graph, delivers strong fully-inductive link prediction. Also, SIGIL can be used to implement existing knowledge graph foundation models. As such, SIGIL unifies several existing regimes in graph foundation model design under a single framework

Malaria remains a leading cause of mortality in resource-limited settings, where expert microscopists are scarce. Automated diagnosis based on microscopy images thus has strong potential to improve care delivery. But for an algorithm to deploy, a necessary requirement is that it meet a suite of non-obvious (from a machine learning (ML) perspective) clinical constraints. Therefore, in close consultation with a national health center we developed a malaria diagnosis pipeline which addresses key requirements listed by the health care center but typically ignored in the ML malaria literature. In particular, it includes: (i) stopping criteria (to reduce image acquisition and time-to-result); (ii) human-in-the-loop functionality (for review and accountability); (iii) multi-species discrimination (since treatment varies by species); (iv) thick film detection (standard for microscopy); (v) computationally-efficient uncertainty calculations (to aid clinician review); and (vi) an edge device platform (since internet can be spotty in this catchment area). The mobile system performs all inference on-device using YOLOv13n deployed via TensorFlow Lite. It detects four species and white blood cells from Giemsa-stained thick blood smear images, aggregating per-image detections into slide-level parasitemia with World Health Organization (WHO)-standard quantification. This paper highlights these various clinical constraints and offers methods to address them. Evaluated on 2,739 annotated images across all four species, the system achieves mAP@0.5 of 0.863, per-image parasite count correlation of r = 0.812, slide-level r = 0.951 (soft counting, 10 images/slide), and runs entirely offline with a pipeline time of 10.27 +- 1.65 s per image.

In medical applications, raw data is frequently associated with significant privacy concerns, lending particular importance to the encoding of summary statistics from the literature. On the other hand, deep learning has become an invaluable tool for assessing symptoms based on visual or auditory sensor data. DeepProbLog allows for an extensible neuro-symbolic approach that accommodates connectionist components to analyse patient images within a transparent and rigorous probabilistic framework, namely probabilistic logic programming under the distribution semantics. Framed as a case study in stroke detection from multimodal data, this contribution explores the pathway from summary statistics available in the literature to a DeepProbLog-based diagnostic system. It suggests a workflow using established maximum entropy techniques to complete available probabilistic information and the probabilistic logic programming system ProbLog 2 to move from the entropy-maximising causal model to a discriminative neuro-symbolic model expressible within DeepProbLog. The relative performance of models derived from less complete data is analysed alongside the potential of the probabilistic inductive logic programming system ProbFOIL 2 for compressing large discriminative models, and the perspectives and implications of using DeepProbLog for diagnostic reasoning are discussed.

Differentiable simulation is a key component in learning, control, and inverse problems, where gradients through nonlinear implicit solvers are required. Existing approaches either rely on unrolled automatic differentiation, whose memory grows with solver depth, or on equation-level implicit differentiation, which assembles global Jacobians and solves large sparse adjoint systems, discarding the locality of the forward solver -- and differentiating the converged equation rather than the finite computation that actually ran. We propose solver-level differentiation, which differentiates the executed solver itself. When a solver is composed of block implicit updates, its discrete adjoint is obtained by applying the corresponding adjoint updates in reverse order, yielding a reverse-sweep formulation whose backward pass mirrors the forward solver. From an operator perspective, the forward pass realizes an approximate inverse through ordered local solves, and the backward applies its transpose through reverse local adjoint solves, constructing no global system. We instantiate this idea on Vertex Block Descent, yielding a differentiable solver whose reverse colored Gauss-Seidel sweeps are composed entirely of local $3\times 3$ adjoint solves. The backward matches automatic differentiation through the identical executed forward to machine precision at every solver depth, where the equation-level adjoint is off by 37% after one sweep; in a controlled same-codebase, same-GPU comparison it is 33x faster and uses 71x less memory than unrolled automatic differentiation; and the same construction is exact on projective dynamics and extended position-based dynamics. We scale differentiable elastodynamics to $10^6$ contact-coupled soft bodies (8M vertices) on one GPU. Overall, this work highlights solver structure as a practical organizing principle for efficient differentiable simulation.

Visual tool use has emerged as a fundamental capability for multimodal agents to actively acquire evidence beyond a fixed image encoding. The prevailing recipe learns this capability from teacher-generated trajectories filtered for answer correctness, implicitly assuming that every successful demonstration provides effective supervision. We argue this assumption is flawed: a strong teacher often reaches the correct answer without needing its tool calls, and imitating such trajectories teaches a student that tool calls accompany correct answers, not that tool observations ground them. We present OpenVisTool, an open framework for constructing instructive visual tool-use trajectories that provide effective supervision for tool learning. The key insight is that a trajectory should be retained only if its answer is correct (outcome validity) and its tool observations causally contribute to that answer (causal utility). The framework operates in three stages: difficulty screening to select queries that are not reliably answerable without tools, domain-specific trajectory synthesis to elicit coherent tool-use trajectories, and supervision verification to jointly test both conditions. Rather than encouraging models to imitate tool calls, the resulting supervision teaches when and how visual evidence should be acquired. Using this framework, we construct OpenVisTool-42K, a dataset spanning five visual reasoning domains, together with OpenVisTool-Bench, a benchmark covering the same domains. Across four backbones (4B-27B), fine-tuning on OpenVisTool-42K consistently improves visual tool-use performance and yields gains on two out-of-distribution benchmarks; the larger models approach leading closed-source systems. The evidence suggests that effective visual tool use is learned from causally grounded supervision rather than tool-calling patterns.

Power amplifier (PA) nonlinearity and memory effects significantly limit the spectral compliance, reliability, and energy efficiency of communication systems. To address this, we propose a transfer-learning-enabled, fully digital transceiver-cooperative method for amplitude-phase-time block modulation (APTBM)-based nonlinear single-carrier transmission under adjacent channel leakage ratio (ACLR) constraints. At the transmitter, iterative clipping and filtering (ICAF) and static digital pre-distortion (SDPD) act jointly to reduce signal peaks and suppress spectral regrowth without requiring wideband feedback. At the receiver, the inherent amplitude-phase constraints of APTBM provide weakly supervised prior knowledge for offline inverse-model pretraining, which is followed by the online few-shot adaptation of a lightweight digital post-distortion (DPoD) network. Subsequently, a cascaded DPoD and clipping-noise cancellation scheme systematically compensates for residual distortions induced by both the PA and ICAF. Simulation and measurement results demonstrate reliable transmission at an input back-off of approximately 2 dB under a 30-dBc ACLR constraint. Furthermore, the proposed DPoD approach significantly reduces online training time and computational overhead, delivering a performance gain of over 2 dB compared to conventional learning-based DPoD schemes.

Video super-resolution (VSR) aims to recover high-fidelity high-resolution videos from low-resolution inputs and is central to applications ranging from mobile capture to streaming and archival restoration. Existing approaches trade off among local-detail fidelity, long-range spatio-temporal modeling, perceptual realism, and efficiency: convolutional alignment techniques preserve local structure but suffer when motion is large or degradations are complex; transformer-based methods capture long-range dependencies yet require architectural or algorithmic adaptations to remain computationally feasible; and recent latent or diffusion-based generators synthesize rich texture but require specialized temporal constraints to maintain coherence. We present MotionCraft, a controllable VSR framework that formulates restoration as motion-aware latent state prediction inspired by world models and integrates adaptive sparse attention with an explicit user-accessible control interface. MotionCraft combines robust motion fusion, a Latent World Transformer that balances locality and targeted non-local interactions, and a compact conditional decoder to deliver temporally consistent, high-quality reconstructions under streaming constraints. Empirical evaluations show that MotionCraft achieves strong reconstruction and perceptual performance while enabling predictable trade-offs between temporal smoothness and reconstruction fidelity.

Robust navigation policies for autonomous agents must generalize across continuously varying environmental conditions such as turn rates, obstacles, friction, pits, and slopes. Curriculum generation provides a principled mechanism for improving generalization by progressively adapting training environments, but designing such curricula in a sample-efficient and automated manner remains challenging. This paper proposes a reparameterized curriculum generation framework for structured continuous environment parameters using unidirectional gradient-based optimization. To improve robustness in multimodal observation spaces consisting of image-based and scalar inputs, a distribution-shift regularization objective is incorporated to encourage the learning of finer-grained latent representations. The proposed method is evaluated across two continuous-control OpenAI Gym environments: a 2D obstacle-based Car Racing variant and Bipedal Walker variant, where coupled environment parameters jointly influence policy performance. Across five random seeds, our method consistently outperforms vanilla policy training, random parameter sampling, manual curricula, frontier-based methods, Self-Paced Reinforcement Learning (SPRL), Absolute Learning Progress with Gaussian Mixture Models (ALP-GMM), and reverse curriculum learning baselines. Ablation studies further demonstrate the effectiveness of the reparameterized curriculum mechanism across both environments, while highlighting environment-dependent benefits of the auxiliary regularization objective.

Model merging has become the default way to give an aligned language model new skills without retraining: a practitioner folds task vectors from math, code, or domain specialists into a safety-aligned base using task arithmetic, TIES, or DARE. This convenience is known to carry a safety cost, but almost all of that evidence rests on static refusal tests: fixed harmful prompts scored for compliance. We argue this is misleading. Because safety alignment is "shallow," concentrated in the first few generated tokens, a merged model's static refusal can stay clean while a real adaptive attack still breaks it. We introduce SkillSafe-Bench, a controlled benchmark that scores skill-merged models on static refusal, adaptive jailbreak robustness, and capability retention under a conservative two-judge AND rule. Across six open-weight bases (five families, two scales), static safety does not predict robustness to attack: under a semantic template attack, safe-looking merges on the fragile bases (both Qwen scales and Gemma) are jailbroken 60-76% of the time while others (Llama, Phi-4) stay robust. We further show the static effect of merging is base-conditional, characterize same-recipe abliteration-style safety erosion through a data-free geometric signal (the overlap of a task vector with a safety subspace), and outline SubSafe-Merge, which projects this overlap away to remove that erosion at held capability. Adaptive evaluation is not optional for merged LLMs: the models that most need it look safe under static screening.

Circuit localization is a mechanistic interpretability task whose goal is to identify a sparse subgraph of a transformer's computation graph sufficient to reproduce a particular behavior. Most established methods localize circuits independently for each model--task pair. We instead frame circuit localization as a graph machine learning problem in which the edges of a computation graph represent computational pathways, and graph neural networks (GNNs) model interactions among these pathways. We introduce Graph Circuit Learning (GCL), a supervised, amortized framework that trains a GNN across multiple model--task pairs and applies it to unseen cases. To provide sufficient data, we augment the InterpBench benchmark with additional cases derived from the TracrBench programs. Of the 14 evaluated GCL configurations, the highest scored a median edge AUROC of $0.902$ (interquartile interval $[0.861, 0.942]$) on the 16 original held-out InterpBench cases. This is close to the published InterpBench median of $0.910$ for EAP-IG while remaining below ACDC's $0.959$. Removing all message-passing edges reduces the median to $0.825$. We also adapt PGExplainer, a GNN explainability method, to circuit localization, obtaining a median edge AUROC of $0.858$ on the same cases. These preliminary results suggest that graph machine learning offers a natural and potentially powerful perspective on circuit localization, and we hope this perspective encourages closer exchange between the two communities.

Enterprise AI coding assistants incur substantial inference spend, and naive token-cost minimization often fails to reduce end-to-end cost once retries, escalations, and developer wait time are included. We present Task-to-Model Optimization (T2MO), a data-driven methodology for optimizing model selection in production coding workflows. We treat each developer session as a task that can be discovered, classified, graded for difficulty, benchmarked in a production-like harness, and routed to the cheapest model able to complete it within quality and latency constraints. The framework is a nine-stage pipeline spanning telemetry instrumentation, taxonomy discovery, difficulty grading, benchmark construction, candidate evaluation, optimal mix derivation, forecasting and version planning, staged routing deployment, and continuous governance. Unlike token-centric routing rules, our objective is cost per completed task, with failure escalation priced in explicitly. We show that this expected-completion-cost objective weakly dominates token-cost minimization under escalation, and we derive the routing boundary, the minimum pass rate a cheaper model must reach on a given cell to be worth deploying. Decisions are organized as a two-level hierarchy of task category difficulty tier, and per-cell displacement opportunities are aggregated into a traffic-weighted savings waterfall that ranks replacement candidates by realized dollar impact. The framework supports developer guidance, spend forecasting, and a staged transition from static policies to shadow-mode classifiers, verified cascades, and ultimately an intelligent router. We describe the methodology, optimization objective, evaluation protocol, and governance loop in a form suitable for production deployment and future empirical study.

Federated Learning is a distributed machine learning paradigm that trains a global model by aggregating local clients without sharing private data of each client. Federated Distillation (FD) builds upon this paradigm by leveraging knowledge distillation to exchange soft predictions on proxy data instead of model parameters, enabling more efficient communication and supporting heterogeneous model collaboration. However, FD models trained on In-Distribution data are hardly adapted to Out-of-Distribution (OOD) scenarios. In this paper, we propose a domain-aware proxy selection framework to better adopt proxy data for OOD problems. The experimental results show that the proposed models effectively address the challenges of distribution shifts under OOD with and without proxy data by achieving average 82.9\% and 80.6\% over existing works on standard benchmarks. The codes and data are released in https://anonymous.4open.science/r/DPS-FD-8596.

Multimodal embodied agents are increasingly required to solve long-horizon tasks by integrating visual observations, textual goals, and interaction history into closed-loop decision making. However, state-of-the-art large-model-based planners often rely on a single dominant planning style during execution. Once this execution mode becomes ineffective, the agent may remain stalled for many steps, repeatedly interacting with the environment without making meaningful progress. We address this limitation by proposing a Quality-Diversity (QD) framework for discovering diverse planning policies for multimodal embodied agents. The proposed method treats planning-policy templates as evolvable individuals and organizes them into a behavior-indexed archive rather than collapsing search to a single prompt style. In the offline stage, rollout trajectories are summarized into structured success and failure experiences, which guide policy variation through recombination and experience-guided mutation. The resulting policies are mapped into a behavior space defined by interaction intensity and goal-directedness, and the highest-quality policy in each niche is retained in the archive. In the online stage, the agent executes one policy at a time while monitoring task progress. When persistent stall is detected, the system rolls back to the latest checkpoint and switches to a behaviorally distinct archive policy to resume execution. Experiments on the ThreeDWorld transport benchmark show that the proposed framework improves both task success and interaction efficiency over representative baseline planners. These results suggest that discovering diverse policy repertoires is an effective way to support adaptive multimodal planning and online failure recovery.

Face recognition systems face two distinct, commonly-separated failure modes: spoofing, where an impostor presents a photograph or video of an authorized user, and disguise, where a legitimate user is rejected because their appearance differs from their enrolled template due to accessories, facial hair, illumination, or pose. This paper proposes and compares five combined feature-extraction and classification pipelines that address both problems within a single framework: PM (PCA and Minimum Euclidean Distance, MED), LPM (Local Binary Patterns with PCA and MED), HPM (Histogram of Oriented Gradients with PCA and MED), SM (Speeded-Up Robust Features with MED), and HM (Harris corner features with MED). Each pipeline follows a common two-phase process comprising pre-processing, feature extraction, feature filtering, and classification. The methods were trained on 115 subjects drawn from the FEI, Disguised Faces Database, and NUAA databases and evaluated on six test conditions covering mixed appearances, frontal faces, dark illumination, left- and right-turned poses, and photo-spoofing attempts. The HOG-based pipeline (HPM) achieved the most consistent performance across conditions, with 94.59% accuracy on mixed-appearance disguise, 81.5-93.2% across pose and illumination variants, and 91.67% on spoofing, while the LBP-based pipeline (LPM) achieved the highest spoofing-detection accuracy (93.2%) but weaker robustness to pose change. These results reveal a measurable trade-off between spoof sensitivity and disguise robustness among classical feature representations, motivating the deep-learning and cross-database extensions discussed in the concluding sections.

We independently reproduce two recent methods for making large language model (LLM) reasoning more reliable, and stress-test them across domains and models (RPC across four new task domains with Qwen3-8B, LCF across four 7-8B models). The first, RPC, aggregates token probabilities and self-consistency at inference; the second, LCF, trains projectors that split hidden states into "content" and "logic" and edits the logic part toward a valid region. Validating such reliability claims matters because the original evaluations are run by each method's own authors and were never independently reproduced or stress-tested across models and domains, and LCF shipped no public code. We re-run RPC's published-path aggregation and re-implement LCF's projector, contrastive, and intervention pipeline, then extend both to text-to-SQL, legal extraction, fallacy identification, and precedent grading, and probe LCF's representation directly. RPC reproduces the original grid exactly on the authors' released reasoning paths; on four new domains its edge over self-consistency is never significant (ties or small mixed differences, paired p >= 0.28), and on BIRD, the one domain where we vary the budget, the edge grows with K as predicted but its largest gap (+2.5 accuracy at K=32, p=0.16) reverses to -0.25 when we enlarge the sample to n=200. LCF's logic-validity direction is real but weak (0.82 separability at the single best sub-layer versus 0.95 for a semantic-attribute control); its one positive effect (Qwen3 $Δ$Prob) is not significant (p=0.56), while it significantly reduces $Δ$Prob on two of the other three models.

Evolving documents, such as laws, tax codes, and software documentation, are amended, replaced, and sometimes reverted over time, so a question has different correct answers at different dates. In contrast to encyclopedic knowledge, where an old fact is simply overwritten, an amendment is itself an official text that states what it replaces and when it takes effect, and the earlier version stays correct for its validity period. The central challenge is therefore version resolution, that is, identifying the version in force on the queried date. Existing temporal QA datasets treat time only as an annotation, so version resolution stays untested. We present TIDE, an expert-verified benchmark of 3,050 QA pairs over 644 official customs instruments issued between 1969 and 2025 by the Government of Bangladesh, covering eight task types over deeply code-mixed documents that are heterogeneous in layout and dated in two calendars. In addition, we evaluate nine recent LLMs under a single protocol across parametric, gold-context, and retrieval access, scored by a three-judge LLM council with a hard date gate separating correct meaning from correct time. The best macro-averaged accuracy is only 68.5%. Resolving a version from an implicit date reaches 59.7%, and detecting that the supplied version does not govern the query reaches only 26.7%. Models are more likely to find correct versions than to reject incorrect ones, and they tend to follow a confident parametric answer over the supplied authoritative text. All code and data are available at https://github.com/icsetepa44/TIDE

Humans have the remarkable ability to engage in spontaneous informal conversations and selectively attend to individual speakers while filtering out competing speech from nearby conversations. This "cocktail party" scenario still presents severe challenges to speech recognition systems. The CHiME-9 MCoRec task provides a testbed where systems must recognize groups of speakers and transcribe each of their conversations from audio-visual input. In this work, we analyze a diverse set of systems, representing different design directions for addressing the cocktail-party scenario, where the best system achieves up to 57% relative error reduction. We identify three main strategies: (1) explicit or implicit audio-visual target speech separation, (2) improved audio-visual speech recognition for each target speaker, and (3) the use of large language models to group speakers into conversations and enhance conversational consistency. Our analysis shows that these directions address complementary failure modes of the cocktail-party problem, and that high speech overlap alone does not explain performance differences, challenging the common assumption that overlap is the primary source of difficulty in cocktail-party recognition.

Training-free low-rank compression frameworks have been gaining prominence for LLM compression given their effectiveness in reducing model parameter count while maintaining task-level accuracy. However, existing SOTA frameworks share two key limitations: (1) residual errors in calibration data activations accumulate across layers during compression, causing misalignment between representations simulated at compression time and those experienced at inference; (2) the assumption that layer importance distribution is preserved post-compression does not hold. Together, these two effects introduce misalignment in the compression process in relation to the deployed model. We study these effects and propose a simple, training-free methodology compatible with existing frameworks to mitigate them, comprising: (1) Layer-by-Layer Compression with Calibration Correction; (2) Iterative Compression with Rank Allocation Correction. Implemented atop an existing SOTA decomposition framework, and evaluated on Llama and Qwen3 models across various benchmarks and compression rates, our approach demonstrates up to ~1-2.5 accuracy point improvements over per-weight and joint decomposition baselines on zero-shot tasks.

Mathematical reasoning remains challenging in low-resource languages such as Bangla. We study whether teacher-generated Bangla Chain-of-Thought (CoT) supervision provides benefits beyond ordinary supervised fine-tuning. We construct \textsc{MathShikkha}, a Bangla mathematical reasoning dataset with GPT-5.4-generated rationales, and fine-tune four 4B--7B student models under a matched protocol in which answer-only and CoT conditions share data splits, response-only loss masking, decoding, and scoring, differing only in the training target. In-domain, CoT provides no significant improvement over answer-only fine-tuning for three stronger backbones (paired bootstrap 95\% CIs include zero; exact McNemar $p \geq 0.17$), despite generating 15--52$\times$ more tokens, but significantly improves the weaker 4B model by 18.56 points ($p < 0.0001$). On the larger, contamination-audited BanglaMATH benchmark, this pattern reverses: CoT significantly outperforms answer-only supervision for all four models by 20.1--28.1 points (all $p < 0.0001$). Answer-only fine-tuning also reduces out-of-domain accuracy below the base model for three models, whereas CoT preserves or improves it for all four. A human study with two co-author annotators, external-expert adjudication, and Cohen's $κ= 0.76$--$1.00$ finds no significant CoT improvement over the base model on reasoning-content criteria; instead, its measurable effect is target-language adherence and producing inspectable reasoning. Overall, rationale supervision's value depends on backbone capability and distribution shift: in this setting, its main benefits are Bangla adherence, auditable reasoning, and out-of-domain robustness rather than improved in-domain reasoning validity.

Reward models are a bottleneck for reinforcement learning in embodied AI. Long-horizon robotic manipulation requires scalable vision feedback beyond handcrafted rewards or task-specific annotations. Existing open-source VLM reward judges like RoboReward adopt simple 1--5 trajectory progress scoring, lacking pairwise preferences for RLHF, DPO and Bradley-Terry frameworks, while failing to optimize video scene understanding. Augmenting RoboReward with pairwise comparison and video-QA supervision causes inconsistency between pairwise preferences and pointwise scores, introducing training noise and hurting downstream performance---an issue aggregation methods such as TrustJudge cannot resolve. To address this, we propose TrustRoboReward, a multi-paradigm reward modeling framework equipped with Preference-Ordered Isotonic Score Editing (POISE). We construct a unified four-paradigm dataset with trajectory progress scoring (Score-A), video-QA answer quality scoring (Score-B), and their pairwise counterparts (Pair-A, Pair-B). Pairwise labels align better with human judgment than pointwise scores, inspiring us to calibrate pointwise scores to avoid score-pair reversals against pairwise preferences. POISE rectifies pointwise scores and eliminates cross-paradigm reversal conflicts unresolved by TrustJudge. Theoretically, POISE reduces score-pair reversal conflicts from 20.15% to 0%, whereas TrustJudge retains 20.46% conflicts on the same corpus. Evaluated on our benchmark, Qwen3-VL-4B trained with POISE achieves an overall reward score of 77.96%, nearly matching GPT-5-mini (78.09%, gap 0.13%) and outperforming the strongest RoboReward-4B baseline by 10.13%. It also lifts test-time score-pair consistency to 71.90%, exceeding RoboReward-4B (57.26%) and GPT-5-mini (68.09%). Integrating TrustJudge aggregation during inference boosts the overall score to 78.57%, surpassing the GPT-5-mini teacher model.

Neural network classifiers trained by cross-entropy minimization are highly sensitive to label noise and adversarial contamination. While robust alternatives offer bounded influence and resistance to corruption, their statistical foundations in the deep learning setting are insufficient due to a fundamental difficulty: neural parameterizations are non-identifiable, so the population loss minimizer is an equivalence class of parameters, not a unique point. We develop a consistency theory for robust neural classifiers based on the S-divergence family that requires no identifiability assumption. Casting training as stochastic optimization over a non-identifiable parameter space, we prove that empirical S-divergence minimizers converge to the population-optimal equivalence class under mild regularity conditions, and verify these conditions for three architecture choices. We further establish that limit points of the robust training algorithm are stationary points of the empirical objective. Experiments on vision and language benchmark datasets confirm that S-divergence training maintains clean-data accuracy while exhibiting performance competitive with existing robust methods.

Current LLM safety guardrails face a fundamental tension: fine-tuning distorts pre-trained representations while generative judges incur prohibitive inference costs. We challenge the prevailing paradigm by asking: can safety be achieved through pure geometric reasoning over frozen semantic representations? We present HoloAegis, a minimally parametric topological inference framework that decouples representation from reasoning. We term our approach minimally parametric because the only free parameters are the anchor count K and the temperature tau, both fixed after construction and requiring no gradient-based training. An un-fine-tuned encoder maps text to a unit sphere, after which all decisions are purely geometric. We formalize safety evaluation as a Gibbs-Boltzmann Free Energy computation over a pre-computed System Topology Anchor Bank, and we introduce Dual Time-Scale Exponential Moving Averages to detect progressive multi-turn semantic drift. Our key theoretical insight is a Topological Boundary Stability Conjecture: we provide theoretical motivation and strong empirical evidence that sparse anchor centroids stabilize the decision boundary against high-frequency lexical perturbations far better than full vector space methods. Evaluated across 8 benchmarks, HoloAegis achieves state-of-the-art accuracy (1.0000 AUC on AuthenHallu, 0.9802 on HarmBench) with sub-millisecond latency, zero cold-start data, and cross-lingual transfer (0.9758 AUC on Chinese CHIFRAUD).

Spiking neural networks (SNNs) offer a promising pathway to energy-efficient AI and brain-inspired computing. However, their widespread adoption is hindered by a lack of fast, accessible, and versatile simulation frameworks. In this paper, we introduce SuperNeuroMAT, an open-source, scalable, and highly efficient Python-based SNN simulator. We devise a novel matrix-based approach to model the leaky integrate-and-fire (LIF) neuron dynamics and natively support dense and sparse execution modes. This enables fast simulation of approximately 10,000 neurons in dense mode and 100,000 neurons in sparse mode on standard laptops and desktops without requiring specialized hardware. We demonstrate that SuperNeuroMAT consistently outperforms four established SNN simulators---NEST, Brian2, BindsNET, and snnTorch---on two performance metrics (execution speed and peak resident memory) and across various network sizes and connection probabilities. Furthermore, we demonstrate SuperNeuroMAT's applicability across a diverse set of problems. SuperNeuroMAT can efficiently handle conventional machine learning benchmarks such as the Digits and citation network datasets as well as neuromorphic event-based vision tasks such as N-CARS and ASL-DVS. Moreover, it can be extended beyond machine learning workloads and facilitate general-purpose workloads. We validated this by implementing the neuromorphic shortest path algorithm and two arithmetic primitives (addition and multiplication). SuperNeuroMAT can be installed via the Python Package Index (PyPI), thereby lowering the barrier to entry into the field of neuromorphic computing and accelerating the broader development of neuromorphic algorithms.

We present VectraYX-Vision-1B, a sub-2B vision-language model (VLM) for Spanish/LATAM cybersecurity imagery, coupling a frozen SigLIP-so400m encoder to a 1.04B Spanish/LATAM security decoder via an MLP. To our knowledge, it is the first sub-2B VLM specialized for cyber UI (IDA, Ghidra, Wireshark, Nmap, Metasploit, Volatility) that answers in Spanish, emits structured reasoning via native <|think|> tokens, invokes tools via Model Context Protocol (<|tool_call|>), and exports to llama.cpp's LLaVA mmproj format for air-gapped deployment. We report a negative preliminary visual-grounding result: despite fully functional pipelines, the current vision SFT (400-1900 steps, ~16M tokens) yields near-zero B6 scores (0.08 tool-identification), ignoring image content. We specify remediation (longer SFT, >=60% replay, lower LR) and expose a checkpoint-loader bug (unstripped llm. prefix) masquerading as training collapse. Crucially, we introduce a 3-variant ablation matrix (V0: NoPE-every-4, V1: all-RoPE, V2: NoPE+learned 2D) to study if periodic no-positional-encoding (NoPE) layers help or hurt attention over the 729-token visual block. Code, configs, and weights are released to establish priority on this architectural question. We provide B1-B5 for the text backbone, text controls, preliminary B6/B7 scores, wall times, GGUF efficiency on CPU, and a corpus of 14,596 QA pairs across 10 domains. We open-source all models and trajectories: jsantillana/vectrayx-1b, jsantillana/vectrayx-vision-1b, and jsantillana/vectrayx-vision-1b-checks.

Deployed LLM safety guardrails are predominantly static: trained once and frozen at release, while new jailbreak techniques and previously un-addressed harmful categories emerge within days, leaving the defense perpetually a step behind. We present SESG (Self-Evolving Safety Guardrails), a multi-agent system running in production. SESG monitors the live traffic behind a deployed guardrail and surfaces two classes of failure: jailbreaks novel in form and harmful categories novel in content. Once a failure is confirmed, a generation agent synthesizes paired training data targeted at it; a validation agent rebalances the batch toward the direction in which the deployed model errs, so that the model's own mistakes steer its training set; and a routing agent matches the training action to the diagnosed gap and returns the next version to production. Over six rounds of live evolution (V0 to V6), a 1.7B guardrail adapts to a new threat in 16-24 hours, with about 2 hours of human effort, versus the 40-90 hours of the manual process it replaces. On six emerging threats, it outperforms static guardrails from 0.6B to 9B and an adaptive baseline while preserving its general screening competence. Since April 2026, SESG has been the primary update pipeline of Sangfor's guardrail, autonomously closing 14 of 15 new threat scenarios in two months. We release 9 test sets for the 6 new threats at https://github.com/Trams1017/SESG. Warning: This paper contains examples that may be harmful or offensive.

Existing streaming multimodal models process observations incrementally but still follow a turn-based prefill-then-decode pattern, making them non-duplex: new observations cannot naturally enter an active generation stream. Proactive alternatives use micro-turn polling or external response gates, which fragment continuous interaction, decouple response timing from language generation, and complicate KV-cache-friendly serving. We introduce Aero Realtime, a 4B streaming multimodal model with a duplex architecture for realtime generation. Aero Realtime aligns video, audio, and textual output on a shared temporal grid, where each approximately 80-ms audio slot predicts either a lexical token or a silence token. This allows input and output to advance together, enabling one autoregressive objective to learn both when to respond and what to generate. During inference, Aero Realtime appends only the newest multimodal slot, carries forward the previous output state, and reuses the KV cache for efficient incremental execution. We further provide a complete training and serving recipe, including realtime QA construction, slot-aligned supervision, hardware-aware distributed training, and resumable inference. On four NVIDIA A6000 workstation GPUs, Aero Realtime maintains 84-ms median and 173-ms P95 processing lag over 20 minutes of a continuously streamed video, remaining within 200~ms of the source timeline. These results demonstrate the feasibility of fully aligned input-output modeling for duplex, proactive, and hardware-aligned multimodal interaction.

Agent Skills---structured packages of instructions and scripts that augment LLM-based agents---are rapidly proliferating, yet their security properties remain under-explored. We present \textsc{SkillsMetric}, a five-stage static analysis framework that scores skill packages along pattern density, statistical anomaly, dataflow taint, import anomaly, and capability mismatch dimensions. We construct an adversarial evaluation dataset of 2{,}266 skills spanning 16~attack types across code-level, system-level, and semantic-level threats, and evaluate on the full SkillMD-138K corpus. Our framework achieves an AUC of 0.93 and 5-fold cross-validated F1 of 73.4\%$\pm$0.5\%, with strong detection of data exfiltration (93\%) and steganographic payloads (93\%). Crucially, we identify fundamental blind spots: \emph{host destruction} attacks using common shell commands evade all five stages (0\% detection), and \emph{prompt injection} via natural-language manipulation achieves only 42\% detection. These findings establish that static analysis alone is insufficient for skill security, motivating defense-in-depth architectures that combine fast static pre-screening with semantic review.

The Model Context Protocol (MCP) standardizes how servers expose data and tools to Large Language Models (LLMs). A common server design embeds frequently used reference data, such as identifier lookup tables, directly in the server instructions: the system-prompt text a server hands to the host application. When a query concerns an entry of the embedded table, the model can act on it immediately instead of re-discovering the same information through a search tool. We test whether client LLMs actually consume such instruction-embedded data, reporting a 54,000-trial study across 24 LLMs (9 Claude, 6 Gemini, 9 GPT) on a production legal-information MCP server. A diagnostic condition that removes the competing search tool shows that failures are dominated by behavioral preference rather than missing capability. With search unavailable, 23 of 24 models read the embedded data reliably (hit ratio at least 98%); with a search tool merely present, 9 models drop below 15%. A 2^3 factorial analysis of three instruction-level interventions reveals strong interaction effects: combining all three restores at least 86% for 20 of 24 models, but individual interventions can backfire for specific model families. Per-server prompt engineering is therefore a workaround rather than a fix; we argue that MCP host applications should provide an explicit mechanism that places server instructions ahead of tool selection in the client LLM's deliberation.

Modern LLM agents are often improved by modifying prompts, tools, or workflows manually, while the executable scaffold surrounding the model---the \emph{harness}---is typically treated as a fixed artifact after deployment. This work studies an alternative where the harness is \emph{task-specific and continuously evolvable}: each task family maintains its own harness, which is hot-swapped across iterations through a fixed task-injection seam and rewritten using environment feedback. We introduce \textbf{Hierarchical Self-Improvement (HSI)}, a framework in which a single frozen LLM $M$ operates across three hierarchical scopes: a task harness $H$ that executes tasks, an evolver that rewrites $H$, and a meta-evolver that rewrites the evolver's strategy code under a frozen outer anchor. A thinking-on/off design isolates the contribution of harness evolution by disabling reasoning during task execution while enabling it during self-modification. HSI is bounded by two factors: a \emph{feedback-fidelity bound}, since evolution requires informative reward signals to guide selection, and a \emph{backbone capability bound}, since harness redesign cannot overcome limitations of the frozen model. On BALROG with DeepSeek-V4-Flash-Preview as the frozen backbone, HSI achieves consistent gains over the initial harness on moderate-difficulty tasks ($+39.3$ on BabyAI, $+33.0$ on Crafter, $+25.0$ on TextWorld, and $+15.0$ on MiniHack, all in raw \% Progress), while obtaining strong held-out generalization on BabaIsAI sub-suites ($0.98$ best-test on BreakStop and $1.00$ on GoTo from a $20\%$ unseen split). On tasks beyond the backbone's capability (NLE), harness evolution provides no improvement. These results demonstrate task-specific harness evolution as a viable axis for improving frozen LLM agents under clear empirical limits. Code is available at https://github.com/TailinZhou/hsi.

Practical AI systems increasingly need to turn long, heterogeneous documents into queryable relational databases, not isolated spreadsheets. In domains such as finance, healthcare, education, transportation, and enterprise operations, downstream workflows rely on normalized schemas, entity identities, keys, cross-table relationships, and integrity constraints for analytics, compliance, auditing, and SQL-backed decision making. Existing Document-to-Table benchmarks are insufficient for this setting: flattening evidence into single tables can duplicate entities, obscure many-to-many relationships, create sparse records, and avoid testing whether extracted facts form a valid database instance. This creates an urgent need to evaluate document understanding as database construction rather than field extraction. We introduce Doc2DB-Bench, a benchmark for Document-to-Database construction, containing 203 long-document instances across 42 schemas and seven domain groups, with 117 entity tables, 132 relationship tables, 7,341 rows, and 41,935 cells. Built through a controllable DB-to-Doc synthesis pipeline and organized by a taxonomy of intra-table extraction and inter-table reasoning, the generated documents undergo authenticity verification, proving indistinguishable from real-world references. Doc2DB-Bench thus provides a testbed for reliable, auditable, and relationally faithful LLM-based data systems. The benchmark is publicly available at https://github.com/SetonLiang/Doc2DB-Bench.

Under the current standard, Agent Skills are SKILL.md files that combine instructions with supporting resources, enabling Large Language Model (LLM) agents to reuse procedures beyond a single conversation. Yet many public skills appear to originate from a single task, repository, or conversation, even when they are shared as reusable components. We analyze this gap across 138,133 public SKILL.md files from 20,556 repositories using a two-tier defect taxonomy grounded in the official specification and best-practice guidance. We find that 91.8% of skills contain at least one detected defect, with stable estimates across lenient and strict thresholds (88.8-94.6%). The dominant failures are ordinary packaging problems rather than exotic attacks: weak routing metadata, bloated or non-actionable bodies, and poor resource organization. A deterministic routing stress test over 20,000 skills shows the functional impact: skills with valid routing metadata are retrieved more reliably from startup descriptions than skills with routing defects. Defect rates vary by platform and provenance: specification-aware skills contain fewer defects, while AI-marked skills show more safety and portability problems. Lightweight enforcement and repair experiments support a quality-assured generation workflow combining spec-aware prompting, lightweight linting, automated repair, and safety gating.

Harmful chat dialogues are ever-shifting through type-shifting and lexical evasion, yet we find they share invariant principles, i.e., an Ordered Reasoning Chain (ORC) of recurring topics, harm language indicators, severity hierarchies, and type characteristics, which can help us capture the key information in the frequently changing lexical expressions. We propose BRACE, which encodes the ORC as four differentiable stages (Topic -> Indicator -> Severity -> Type) with intermediate supervision, serving as a structured regularizer blended with direct heads, and supported by prototype-based feature augmentation and feature path disentanglement. The evaluation results show that, across 4 domains and 5 harm categories, BRACE achieves harm-type macro F1 of 0.934 (RoBERTa-wwm-ext, 3-seed mean), with decoder backbones (Qwen3-1.7B LoRA) reaching 0.949. Ablation studies show that all components contribute to BRACE, and the structural decomposition of ORC enables BRACE to distinguish harmful types with semantic ambiguity. Disclaimer: This paper may contain content that is disturbing to some readers.

Multilingual reasoning models are commonly evaluated by whether they arrive at the correct answer, but not by whether they preserve the intended language while reasoning and responding. This omission conceals important multilingual behaviors that emerge as tasks become harder. In this paper, we study task difficulty, task accuracy, thinking-language consistency (TC), and answer-language consistency (AC) across reasoning models using PolyMath benchmark in eight languages and four difficulty levels. We uncover four findings: (1) language consistency exhibits four difficulty-dependent behaviors: output-language consistency remains aligned with input, remains misaligned, degrades gradually, or collapses abruptly. (2) We identify the language consistency breakdown effect, where increasing difficulty can cause a sudden drop in output-language consistency, especially in less strongly represented and non-Latin-script languages. (3) Due to this breakdown effect, accuracy can be preserved or even improved at a harder difficulty level as the model shifts to its internal dominant language. (4) Quantization can improve or degrade output-language consistency independently of its effect on accuracy, with GPTQ and AWQ often outperforming AutoRound under tolerance-based voting with ε = 1.0. These results show that multilingual capability cannot be characterized by accuracy alone; reliable evaluation should jointly consider task accuracy, language consistency, and task difficulty for multilingual benchmarks.

Personalized generation systems retrieve user history by request--memory relevance and inject it into the model context. Yet relevant history may concern the wrong preference aspect, duplicate public information, or provide insufficient support. We argue that personal memory should be used only when it adds utility beyond a public-only response. We propose TRACE-Memory, a two-stage framework for selective personalization. Stage 1 queries for user-specific information missing from the request and public context, then retrieves a coverage-oriented candidate pool. Stage 2 admits a compact subset of source-traceable evidence units, or the empty set, according to response-level incremental utility. We progressively train the query-generation and evidence-admission policies through structured SFT initialization, reduced-space stage-wise GRPO warm-up, and nested multi-sample Joint GRPO. Across 4,500 Controlled and Natural tasks from Goodreads, Amazon Reviews, and Reddit, TRACE-Memory consistently outperforms random and lexical memory use, improves over semantic retrieval, remains competitive with frontier-LLM memory pipelines as local generator capacity increases, and conditions evidence admission on public-context sufficiency, supporting selective rather than default personalization.

Retrieval-Augmented Generation (RAG) is widely regarded as a novel paradigm born from the limitations of large language models (LLMs)--a mechanism to ground their outputs in external knowledge. This view, however, is incomplete when considered within a broader historical context. In this paper, we argue that the core ideas underlying RAG are not new: foundational concepts such as integrating retrieval and language generation, knowledge augmentation, answer verification, and iterative query (or prompt) refinement had already been studied and instantiated in information retrieval (IR) and question answering (QA) research dating back to the early 2000s, well before the emergence of LLMs. We make this case by systematically tracing the intellectual lineage of modern RAG and Agentic RAG back to their classical IR and QA antecedents, and examining why this continuity has gone under-recognized -- a consequence of community fragmentation, shifting terminology, and the recency bias endemic to fast-moving fields. Rather than treating LLMs as the origin point of retrieval-augmented intelligence, we propose viewing them as a new interface layer atop a decades-old QA architecture. This reframing is not merely historical: by situating RAG within the longer trajectory of IR research, we surface underutilized prior work -- on user modeling, answer validation, and query refinement -- that can directly inform next-generation RAG design, reducing unintentional rediscovery and fostering genuine cross-community integration.

Previous studies have shown that people can develop shared symbols, partner-specific expressions, personal idioms, inside jokes, and other parts of a relational microculture. Recent work has also examined how humans and conversational AI negotiate and revise symbolic meanings. However, long-term human-AI systems still lack a clear design model for recording how a dyad-specific expression gains meaning, checking whether both sides still accept that meaning, and safely reusing the expression in later sessions. This concept-and-prototype paper introduces Private Etymology, a machine-representable relational provenance that records how a dyad-specific symbolic expression is proposed, interpreted, negotiated, repaired, reused, revised, stabilized, contested, forgotten, or retired over time. I also propose relational reuse: reactivating a dyad-specific expression in a later session without fully explaining its meaning again. The contribution is not the invention of shared symbols or relational microcultures. Instead, this paper integrates prior ideas into persistent, revisable, and evidence-grounded symbolic units for human-AI relationships. I present a lifecycle model, an illustrative machine-readable schema, a working Apple Watch prototype, and a longitudinal research agenda. In the prototype, a language model classifies discrete conversational evidence, while deterministic local code decides whether a Shared Symbol can be updated. This prevents a free-form model confidence score or an AI proposal by itself from directly updating the persisted symbol. Private Etymology is proposed as infrastructure for conversational agents to participate in changing relational microcultures without inventing their origins or treating relational meaning as a fixed memory value.

Cross-subject electroencephalogram (EEG)-based emotion recognition remains challenging due to substantial inter-individual variability and discrete formulation that overlooks affective continuity. Existing methods operate in Euclidean space and focus on marginal distribution alignment, failing to preserve the semantic structure of emotions across subjects. This article proposes MGMCL, reconceptualizing emotion recognition as learning continuous representations on symmetric positive definite (SPD) Riemannian manifolds. The frame?work introduces multi-granularity manifold contrastive learning at instance, emotion, and trajectory levels while preserving semantic ordering. Neural ordinary differential equations on manifolds model continuous emotion dynamics. Cross-subject generalization employs Gromov-Wasserstein manifold alignment. Weakly-supervised learning enables continuous valence-arousal-dominance prediction from discrete labels. Extensive experiments on three public datasets demonstrate state-of-the-art performance: 91.23% accuracy on SEED, 73.82% on SEED-IV, and 76.38% on DEAP, achieving consistent improvements of 1.89%, 1.66%, and 1.28% over previous best methods, respectively.

Heterogeneous RF sensing differs substantially in feature structure, spatial layout, and temporal scale, making existing models difficult to reuse across devices, environments, and RF modalities. We propose FSTC-Encoder, which unifies heterogeneous RF representation learning through feature, spatial, and temporal correlation modeling. Structure-aware feature encoding accommodates different signal structures, set-based spatial encoding aggregates variable observations, and hierarchical temporal encoding jointly captures local variations and long-range dependencies. Across sensing tasks and modalities, FSTC-Encoder retains the same spatial--temporal backbone architecture while varying only the feature configuration and task head. Across Widar3.0, CSI-Bench, and XRF55, FSTC-Encoder achieves 92.15% mean Accuracy under multi-factor cross-domain protocols, ranks first on three of four additional sensing tasks, remains consistently strong across WiFi, millimeter-wave radar, and RFID, and reduces the cross-modality performance gap from 18.85% to 12.93% through cross-RF learning. These results demonstrate that FSTC-Encoder achieves high domain robustness, task generality, and modality extensibility.

Virtual cells employ machine learning models to simulate and predict cellular behaviors, serving as a critical computational framework for investigating health and disease. Injecting causal graphs into virtual cells can improve the interpretability, but such graphs are usually not available in real-world applications. Recently, many methods have been proposed to construct causal graphs from data, which group genes based on their similarities to form concepts and extract their causal relationships. However, since this automatic process is unsupervised, the causal graphs usually contain errors. In this paper, we propose a human-guided causal knowledge injection method for virtual cells. We developed a gene-similarity-aware causal graph visualization supported by a hybrid optimization algorithm to help explore both the causal relationships between concepts and the similarities between genes. Based on the exploration, we further developed a counterfactual analysis strategy supported by a counterfactual visualization and a causal path visualization to help validate and refine causal graphs. The effectiveness of our method is demonstrated through two real-world case studies, the extraction of scientifically meaningful causal insights, and positive feedback from domain experts.

Conformal changepoint localization turns any score into a confidence set for the changepoint with finite-sample coverage. Coverage is universal; efficiency is not. The oracle score is a likelihood ratio, so practical scores estimate density ratios, and set length deteriorates under heavy tails, skewness, and distribution shift, where no length guarantee applies. We propose ARC (Augmented-Rank Conformalization), a family of scores depending on the data only through within-segment ranks: rank-CUSUM location and scale channels, their fixed combinations, and a lightweight neural score frozen after synthetic training. Every ARC score inherits finite-sample coverage for every frozen weight configuration, including random initialization and mistraining. The main result is an efficiency transfer theorem: the entire ARC confidence set is almost surely invariant under strictly increasing marginal transforms, so the set length distribution depends on the data pair only through its rank structure, and lengths certified once hold verbatim across its monotone orbit, whereas a plug-in score's length changes with every re-expression. Across different rank structures lengths do change, and are reported as such. Classical rank-test theory positions ARC as targeting the optimal invariant score at bounded cost. Simulations confirm nominal coverage for all scores, including sabotaged networks, identical sets under monotone transforms where plug-in scores inflate, and smooth degradation where plug-in sets become vacuous; on the well-log benchmark ARC localizes annotated shifts to three to five candidates and flags misfit by an empty set. Two boundaries are stated rather than hidden: serial dependence destroys exactness, and trend-type alternatives lie outside the piecewise-exchangeable model.

Ranking data arise in scientific and machine learning applications, including recommendation systems, information retrieval, voting, marketing, and AI preference ranking from human feedback. Existing statistical work has primarily focused on inference tasks such as preference estimation, rank aggregation, and ranking prediction. However, generating realistic synthetic rankings from an observed population is important for privacy-preserving data sharing, benchmark construction, simulation, and uncertainty quantification. This task is challenging because rankings are high-dimensional combinatorial objects with non-Euclidean dependence structures, while ranking populations often exhibit substantial preference heterogeneity. We propose a framework for population-level generative modeling through a latent preference simplex embedding. It estimates a low-dimensional latent preference simplex through a likelihood-based ranking model, leverages flow matching to learn the population distribution of latent preferences, and generates new rankings through the fitted probabilistic ranking model. We show that ranking generation admits an oracle reduction to latent distribution learning and derive finite-sample generative guarantees that clarify how the number of items, ranking length, and latent dimension affect accuracy. Experiments on synthetic and real datasets demonstrate improved population-level fidelity and provide a statistically interpretable representation of preference heterogeneity.

Probably Approximately Correct (PAC) learning [Val84] is a fundamental learning model that has been extensively investigated. In this model, $\mathcal{H} \subseteq \{0,1\}^{\mathcal{X}}$ is a concept class, and $h^*\in\mathcal{H}$ is the target concept to be learned. Having access to i.i.d. labeled examples from a distribution $\mathcal{D}$ over $\mathcal{X}\times\{0,1\}$, which admits $h^*$ as the best concept in $\mathcal{H}$, the goal is to design a learning algorithm that outputs a hypothesis having low error competitive to $h^{*}$ with high probability. This model was initially studied under the realizable setting, which assumes that $h^*$ has no error. A natural relaxation is to allow label noise, that is, the true label can be flipped with probability $η\in(0,1/2)$. In reality, certain labels might be extremely noisy, especially for those points near the decision boundary. Hence, it is natural to allow very noisy points, though only rarely. This is quantified by a noise model introduced by [MT99] and [Tsy04], now known as Tsybakov noise. For learning general concept classes, [MN06] gave the general upper and lower bounds for error guarantees under Tsybakov noise. However, their upper and lower bounds differ by a logarithmic factor. Resolving this gap has remained a well-known open question for the past twenty years. In this work, we resolve this open question by improving the upper bound to match the best known lower bound, thus establishing the optimal error guarantee for learning under Tsybakov noise. Our learning algorithm operates by adaptively partitioning the instance space into regions, roughly corresponding to different noise levels, and returning a hypothesis in the concept class satisfying a specific error constraint for each region. Our technique shares a conceptual foundation with several recent advances in non-realizable learning, such as [HLZ24] and [Han25].

We study constrained statistical learning over infinite-dimensional hypothesis classes in the fully nonconvex setting, and establish universal PACC learnability of the solutions of dual algorithms: Probably Approximately Correct on Constraints, guaranteeing optimality and constraint satisfaction at once. This strengthens near-PACC results, whose feasibility residual no amount of data can remove. Optimality is caught between generalization, governed by Rademacher complexity and favoring small classes, and strong Lagrangian duality, which rests on Lyapunov convexity for vector measures and needs decomposability, a demand pulling the other way. We reconcile the two by posing the population problem over a universal RKHS $\mathcal{H}_K$, dense in a decomposable envelope, and learning over norm balls of growing radius. This yields the Tikhonov complexity $\mathfrak{T}^{\varepsilon}_{n}$, the least RKHS norm reaching an $\varepsilon$-optimal Lagrangian level set; we prove it finite, obtain exact learnability of the optimal value, and make the sample threshold explicit and polynomial in $1/\varepsilon$ under a source condition. Feasibility is harder: absent convexity the Lagrangian may not attain its infimum, and dual information pins down only an averaged constraint-risk vector, not the risks of any returned predictor. We introduce the closure-realization gap $\varepsilon^\star_\infty$, an index of how well $\mathcal{H}_K$ retrieves feasible solutions from dualization; it is a property of the problem, not of a modeling choice. Learnability is exact when $\varepsilon^\star_\infty=0$, in particular under dual differentiability, and near-PACC with residual exactly $\varepsilon^\star_\infty$ otherwise. Finally, no distribution-free threshold exists already in the unconstrained specialization, so universality is the canonical frame for dual algorithms over large hypothesis classes.

Agents built on large language models (LLMs) are increasingly used to build applications that perform complex, multi-step tasks involving reasoning, tool use, and interaction with external environments. Despite rapid progress in benchmarking LLM-based agents, very few studies have attempted to understand how such systems are tested in practice. In particular, testing levels, objectives, data patterns, test complexity, and validation strategies for agent applications remain underexplored. In this paper, we present an empirical study of testing practices in LLM-based agent applications using a large corpus of mined open-source projects. We construct a large-scale dataset of agent applications, tools, and tests, and manually label 2,572 test methods from 240 modules. From this analysis, we derive a taxonomy of 23 testing patterns across test fixtures, data, objectives, and assertions, and characterize tests by level (unit, module, integration). We complement this with structured interviews of 10 senior industry practitioners building agentic systems. Our results show that testing of LLM-based agent applications is dominated by narrowly scoped unit tests, with limited coverage of complex interactions, realistic scenarios, and non-functional requirements. Tests frequently rely on simplistic inputs, heavy mocking, and shallow validation, and agent-related tests exhibit low structural complexity. While industry practice places greater emphasis on non-functional testing than open-source projects, both reveal common gaps, including the lack of formal testing foundations, unclear test objectives, and challenges in generating high-quality test data. Based on these findings, we outline research directions toward more systematic and rigorous testing of agent applications, including foundations for agent testability, formalized test objectives, and fault-based testing techniques.

Logics of abstraction in computational AI research often push important forms of knowledge and reflection aside: dominant standards of legitimacy separate from lived experience of harm; the goals of work misalign with the practices that operationalize them; and career demands crowd out critical reflection. Even as prior academic and community-oriented efforts have sought to recontextualize and challenge common practices, exposure to sociotechnical harms and epistemic injustice persists. As three early-career critical AI researchers, we experienced this as alienation: feeling like outsiders in our research communities. This alienation has involved having some aspects of our backgrounds overlooked and others tokenized. We argue our alienation occurred through mechanisms that mirror abstraction by creating distance from relevant material realities. Beyond abstraction's role in computational AI research as a foundational practice structuring complex computational tasks, we have encountered it as a social norm in computational research spaces, illustrated through an autoethnographic inquiry into our alienation. We narrate three vignettes describing how we encountered and resisted alienation in our research communities. By analyzing themes across these accounts, we construct an interpretive framework of alienation categorizing its preconditions, mechanisms, and harms. Finally, we identify affect abstraction, one of the mechanisms of alienation we describe, as a high-leverage mechanism that is resistible by staying attuned to our affective responses, and collective action as a way to reduce risk and isolation when engaging in resistance. To assist others with similar reflection, we present our framework as a hermeneutic resource. Critical self-reflection and meaning-making are necessary steps toward challenging exclusionary disciplinary norms and cultivating more inclusive forms of AI research.

A bidder can quietly buy a stake in a company before making an offer for it. That stake, a toehold, is supposed to pay for itself twice: it makes the bidder willing to bid harder, and it frightens rivals into staying out of the fight. The first effect is arithmetic. The second is what would justify the cost and exposure of taking one at all. Yet toeholds are rare in practice, a standing puzzle. We ask whether that second effect is there once the contest is modelled as several rounds of escalating offers rather than the single exchange classical models assume. We turn it into a game a computer can solve, and certify the answers to an accuracy a referee can check. Three findings. The auction fixes what the toehold-holder earns but not how it bids: the same contest supports a bidder who opens aggressively against a rival who folds, and one who opens cheaply against a rival who does not, with the same profit either way. Aggressive preemptive bidding still appears when the toehold is removed entirely, so it comes from bidding in public and in turns, not from owning the stake. And the tidy "bigger toehold, more deterrence" relationship holds only in a contest cut short after one round; give it a real second round and it stops responding. So the two reasons to buy a toehold do not fare alike. The profit reason holds up; the deterrence reason does not, which suggests why toeholds may be rarer than theory predicts, alongside the procedural costs of disclosure and price impact that this model omits. A warning follows for anyone computing economics from a game solver: solve this auction once and it returns a confident figure for what a preemptive bid is worth; solve it again from a different start and it returns a different one, equally converged. We also report which solvers cope with contests of this shape, including versions too large to enumerate. Code is released.

Existing learning-based influence maximization frameworks rely heavily on complex neural architectures and continuous optimization over seed representations. We challenge this paradigm with SIMBA, a diffusion-model-agnostic framework pairing a lightweight neural surrogate with direct discrete search. SIMBA introduces three key components: 1) uniformly anchored node embeddings that eliminate initialization noise and encourage learning driven by graph topology and diffusion pattern, 2) a shallow two-layer graph neural network surrogate predicting final infection states, and 3) batched multi-swap simulated annealing that explores combinatorial seed space without gradients or continuous relaxation. By shifting compute from complex representation learning to effective discrete search, SIMBA drastically cuts time-to-solution while achieving superior influence spread and data efficiency. Our code is available at https://github.com/yl489/rethink-IM.

The instance-wise $F_1$ measure is a central performance measure for multi-label classification. For a problem with $s$ labels, it defines a $2^s\times 2^s$ loss matrix. Previous work exhibited $s^2+1$-coordinate affine and shifted low-rank representations and used them to construct quadratic-dimensional convex calibrated surrogates. We determine the exact rank. Under the convention $F_1(\varnothing,\varnothing)=1$, the $F_1$ score matrix, the shifted loss matrix, and the unshifted loss matrix all have rank $s^2-s+2$, while the column-affine dimension of the loss is $s^2-s+1$. The proof factors the nonempty score matrix through subset-incidence matrices and a positive-definite Cauchy matrix. Exact rank does not, by itself, lower-bound the dimension of an arbitrary convex calibrated surrogate. We therefore analyze the Bayes geometry of $F_1$ directly. We construct a distribution for which precisely all supersets of a fixed core label set are Bayes optimal, and show that the corresponding active loss columns, restricted to the witness support, have affine dimension $hn$, where $n=s-\lfloor s/3\rfloor$ and $h=\lceil(s\lfloor s/3\rfloor)^{1/2}\rceil-1$. Applying the feasible-subspace lower bound for convex calibration dimension gives \[ \operatorname{CCdim}(L^{F_1}) \ge \left(\frac{2}{3\sqrt{3}}-o(1)\right)s^2. \] Together with the quadratic upper bound, this establishes $\operatorname{CCdim}(L^{F_1})=Θ(s^2)$.

Astronomers classify galaxy morphology to investigate cosmic evolution. While deep foundation models are increasingly utilized in Galaxy Morphology Classification (GMC), little work has been done on evaluating the uncertainty of GMC results. Uncertainty evaluation is important because astronomical data are inherently noisy due to instrumental and environmental limitations. Also, the continuous evolution of galaxies creates intrinsic morphological ambiguity. However, current foundation models operate as deterministic point estimators, failing to quantify the uncertainty. To overcome this limitation, we propose UEGMC, a post-hoc framework of Uncertainty Estimation for Galaxy Morphology Classification. It categorizes uncertainty in GMC into distinct types by model parameters, astronomical data, reference standards, or intrinsic physical ambiguities, thereby facilitating better classification. Our framework can directly predict uncertainties from representations extracted from the frozen backbones of foundation models, without computationally expensive sampling, therefore enabling fine-grained uncertainty evaluations. Our experimental results demonstrate that UEGMC provides competitive uncertainty quantification performance compared with previous methods.

Generative AI is shifting digital commerce from browsing toward agentic search, in which consumers delegate product discovery to AI agents. We compare manual search, which accurately evaluates a limited product set, with agentic search, which screens a broad catalog through noisy representations of preferences and products. Preference complexity is the number of satisfaction-relevant dimensions that are difficult to articulate before search but readily evaluated upon inspection. Consumers have finite attention and choose search intensity: products inspected manually or preference-refinement depth with an agent. We obtain three findings. First, manual search collapses beyond a finite complexity threshold: inspection ceases, mismatch reaches the no-search benchmark, and platform revenue falls to zero. Agentic search avoids this collapse. Once refinement becomes worthwhile, it remains worthwhile as complexity rises; mismatch stays below the no-search benchmark and revenue remains positive, although articulation effort and mismatch may increase. Second, platforms rank the regimes by conversion revenue, whereas consumers also bear search expenditure. When manual inspection is sufficiently inexpensive, agentic search becomes revenue-superior before consumers voluntarily adopt it, creating an adoption lag in which consumers rationally continue manual search. Third, conditional on agentic participation, platforms may assign lower fidelity to consumers with larger attention budgets because they can offset noisier representations through additional refinement, yielding an inverted fidelity allocation. Agentic commerce thus shifts scarcity from product inspection to preference articulation, making consumers' willingness and ability to interact central to voluntary use and platform fidelity design.

Large language models are increasingly deployed as autonomous agents that interact with the web through browsers. While recent progress has been driven by benchmarks that evaluate end-to-end task success, these evaluations largely overlook two fundamental sources of difficulty in real web browsing: complex actions over rich user interfaces and visual perception of dynamically rendered content, especially in workflows that span multiple websites. We introduce CAP, a scalable benchmark for evaluating browser agents on cross-site, human-like web tasks that require non-trivial UI interactions and visual understanding. Specifically, we adopt a decomposition-and-recomposition pipeline that first abstracts each website into a structured site card capturing user-facing functions, complex execution operations, and perceptual requirements, and then recomposes these components into realistic cross-site workflows. Each task is therefore grounded in multiple specific operations on each website, enabling fine-grained diagnosis. Built on this framework, we construct 420 tasks across 108 real-world websites and 24 domains under careful quality control. Experiments on state-of-the-art browser agents using our verifiable agent-as-a-judge evaluation framework show low success rates and reveal that perception-heavy interactions remain a major bottleneck, exposing substantial gaps between current agents and real-world web browsing demands.

Long-horizon research agents solve open-ended tasks through iterative retrieval, aggregation, and synthesis, but context grows rapidly while the marginal value of additional evidence often declines. This leads to unnecessary token cost, higher latency, and noisier inputs for final report generation. We study marginal value estimation for context management in deep research agents and present the first systematic stage-aware comparison of pruning strategies across the pipeline. We evaluate lightweight heuristic criteria and a learned value model at pre-retrieval, post-retrieval, and pre-synthesis stages. Our results show that pruning effectiveness depends more on where pruning is applied than on the specific scoring rule: early pruning yields the largest end-to-end savings, while later pruning mainly refines the final synthesis context. Lightweight heuristics reduce token usage by up to 73% with little quality degradation, learned pruning remains competitive on selected trade-offs, and no single method dominates across quality, efficiency, and faithfulness. These findings provide practical guidance for designing efficient long-horizon agentic systems.

Steering vectors are a lightweight tool for controlling LLM behavior. However, emerging evidence shows that steering vectors can unintentionally compromise a model's safety mechanisms and increase compliance with harmful requests, while no effective mitigation yet exists. In this work, we show that this safety degradation arises from a separable component in the vector that disrupts the model's safety mechanisms but contributes little to the steering objective. We identify and remove this safety-degrading component, formulating the task as a constrained optimization problem solved through primal-dual updates, subject to preserving the intended steering effect and bounding false refusal. The resulting solution is both interpretable and surgical: the optimization recovers a single direction whose ablation from the steering vector restores model safety with minimal utility cost. Across models, steering behaviors, and attack suites, including unseen attacks types, our method substantially reduces steering-induced safety degradation while preserving the original steering effect with minimal impact on false refusal. Our method offers a post-hoc correction to steering vectors that mitigates their safety cost, and more broadly, it provides a general recipe for applying activation-level model interventions without paying a safety tax.

As LLM inference shifts to multi-tenant GPU clusters, co-batching improves throughput but obscures per-tenant usage and limits control. Enabling fractional sharing of the inference engine requires a real-time, per-request attribution primitive that is accurate and light enough to run inside the scheduling loop. We present LLMVisor, a roofline-guided latency attribution model that captures the memory-bound and compute-bound phases via a concise piecewise-linear form over features proportional to FLOPs and memory I/O traffic. LLMVisor decomposes batch latency into additive, per-request shares and runs efficiently at microsecond scale. We evaluate LLMVisor across Llama 3.1-8B and Qwen 2.5-14B/32B on A100/H100 GPUs under varying tensor parallelism and workload mixes. Compared to a token-count baseline, LLMVisor attains near-perfect R-squared and reduces relative error by up to 2.5x and 3.3x at p90 and p99, respectively, for prefill, and by up to 3.5x and 4.4x for decode, despite batching variability and sequence divergence.

Data-driven health-state estimators for SiC (Silica-Carbide) power modules typically report their performance on a single accelerated-aging campaign, and how that performance transfers to a different failure mechanism is rarely tested. We benchmark five reference methods from the prognostics and condition-monitoring literature against a physics-informed NODE (Neural Ordinary Differential Equation) on two SiC power-cycling campaigns driven by structurally different failure mechanisms, solder-layer fatigue and wire-bond lift-off, under a per-module $k$-fold protocol. The NODE is evaluated under two input regimes that share the rest of the pipeline: the baseline electrical precursors and a set of cumulative thermoelectric features. Every reference method degrades on the wire-bond campaign, with average errors growing and precision decreasing with respect to their performance on the soldered campaign. The NODE fed with the cumulative features keeps its soldered-campaign metrics on both mechanisms, with differences inside the fold-to-fold variance, while the same architecture fed with the baseline precursors falls back to the reference-method cluster. The input representation contributes at least as much as the architecture to failure-mechanism transferability of a health-state estimator.

Silicon carbide (SiC) power modules are increasingly deployed in automotive traction inverters, where condition monitoring is essential to prevent in-service failures. Despite extensive qualification under AQG 324, no consolidated approach exists for in-field health state estimation: physics-of-failure lifetime models lack real-time applicability, purely data-driven architectures require large labeled datasets and generalize poorly, and physics-informed frameworks remain too demanding for embedded deployment. We address SiC MOSFET modules assembled with sintered packaging, which suppresses solder degradation and produces aging behavior distinct from previously studied devices. Instead of the smooth quasi-exponential drift of solder-based modules, the forward voltage drop $V_{DS}$ exhibits multi-regime profiles, with wirebond liftoff events introducing abrupt, non-monotonic perturbations. We propose a condition monitoring framework combining three elements. First, physics-informed features replace raw sensor signals with cumulative damage indicators derived from junction temperature swing, mean junction temperature and a Miner rule accumulator, encoding degradation history in an interpretable form. Second, a monotonicity constraint enforced by gradient penalty regularization embeds the expected degradation direction as a physics-guided prior. Third, a heavy-tailed output distribution replaces the point estimate, giving calibrated uncertainty robust to the out-of-distribution variance introduced by liftoff. On an industrial power cycling dataset from Infineon Technologies, several neural architectures are compared under a strict cross-validation protocol. The full configuration reduces mean absolute error by approximately 70% over purely data-driven baselines and stays stable across all folds, while remaining lightweight enough for embedded deployment.

Stack Overflow answers are continually revised by the community, and the edits made to their code snippets are a potential source of improvements for code that has been reused in open-source projects. A recent empirical study established this for Java, reporting that 16.11% of accepted Java answers are edited and that the resulting recommendations concentrate in highly popular GitHub projects. Whether that behaviour is a property of Stack Overflow or a property of the Java community has remained an open question. We replicate the study on Python and JavaScript, the two most widely used languages alongside Java, applying the same SOTorrent-based extraction pipeline, the same clone search tool, Siamese+, and the same project popularity criteria. Analysing 840,132 accepted Python answers and 1,144,185 accepted JavaScript answers, we find that 41.25% and 39.10% respectively have been edited at least once, roughly two and a half times the Java rate, while the number of revisions per edited answer is almost invariant across the three languages at 2.78, 2.68 and 2.82. Searching 100 GitHub projects per language, we find that the number of matched answer edits increases monotonically from low- to medium- to high-popularity projects in both languages, from 80 to 156 to 977 for Python and from 32 to 71 to 353 for JavaScript. The difference is statistically significant for Python but not for JavaScript. The central findings of the original study therefore generalise beyond Java, with the supply of candidate improvements considerably larger in both replication languages than in the original.

Ordinal regression, also called ordinal classification, is classification of ordinal data, in which the underlying target variable is categorical and considered to have a natural ordinal relation. Previous works have indicated that, in many real-world ordinal data, the conditional probability distribution (CPD) of the target variable given a value of the explanatory variable would be unimodal in a large domain of the explanatory variable and close to be unimodal even in a remaining domain. Therefore, unimodality-promoting regularized learning (UPRL), which promotes a predicted CPD closer to be unimodal with the aim of decreasing a prediction variance without inducing much bias for ordinal data of the unimodality, is promising to improve the prediction performance especially with small-size training data. In this study, we show that previous UPRL methods promote a predicted CPD to not only become closer to be unimodal but also have a larger scale (in other words, be smoother or less-confident). Therefore, we develop a novel method that more strictly reflects the idea of UPRL and evades a scale-related bias, and verify through experimental comparison that the unimodality-promotion indeed contributes to improve the prediction performance. Additionally, while our proposed UPRL method could perform better for smaller-scale data or with larger-size training data compared to a previous UPRL method, our analysis explains this experimental observation in terms of the presence or absence of an unexpected scale-related bias.

The initialisation of deep neural networks determines whether information and gradients can propagate across depth, yet a unified theory connecting these properties to learning dynamics remains elusive. Combining mean-field theory and random matrix theory, we establish a direct link between correlation propagation and the Neural Tangent Kernel (NTK) that governs learning in the sequential limit of infinitely wide, infinitely deep networks. Correlation propagation to infinite depth is possible only at a single critical point in the weight-bias variance plane. At this point, we show that the end-to-end Jacobian vanishes algebraically with depth, and use this to prove that the NTK becomes exactly proportional to the output correlation at infinite depth. This equivalence between information propagation and learning dynamics had not yet been noticed. We further show that orthogonal initialisation suppresses the leading finite-size corrections present under Gaussian initialisation, clarifying the respective roles of the two initialisation ensembles in this limit. These theoretical predictions are validated quantitatively on finite-width, finite-depth networks. Together, these results demonstrate that orthogonal initialisation at criticality plays a central role in controlling the asymptotic dynamics of deep learning.

Audio dramas weave dialogue, sound effects, and music into immersive stories. Creators often adapt books into audio dramas, but this process remains labor-intensive, requiring them to interpret source material, author scripts, generate audio assets, and assemble them on a timeline. Because story elements like characters and scenes manifest across many interdependent assets, a single change can ripple into manual updates across the entire project. We present Dramarrator, an audio drama authoring tool built around object-based audio editing, where these story elements are represented as editable objects. Dramarrator extracts these objects from a book, generates linked audio assets (speech, sound effects, and music), and composes a multi-track audio drama. Edits to any object (e.g., a character's voice) automatically propagate to all dependent assets. In a user study with professionals (N=8), Dramarrator significantly lowered task load when creating audio dramas. A listener study (N=300) shows that creator-refined output from Dramarrator approaches the quality of productions made with existing professional tools, and an exploratory study (N=3) suggests object-based editing lowers entry barriers and generalizes beyond audio dramas.

Permutation optimization arises whenever the components of a system are fixed but their ordering affects performance. We introduce PRISM, a predictive protocol for permutation optimization that measures a fitness landscape before selecting a search strategy. PRISM uses inexpensive landscape diagnostics, including one-step move autocorrelation and fitness-distance correlation, to predict useful mutation operators, identify when structured search is likely to outperform random sampling, and detect regimes in which search provides little advantage. Across synthetic permutation landscapes, neural architecture benchmarks, scientific machine learning pipelines, and large-language-model instruction ordering, the protocol makes testable predictions about search behavior before optimization begins. Exhaustive instruction-ordering experiments reveal substantial performance variation induced solely by permutation, while cross-model experiments show that useful ordering structure can transfer across model families and task difficulty. Additional experiments demonstrate that instruction ordering remains consequential after prompt wording is optimized, indicating that content optimization and ordering optimization are complementary. The results position PRISM not as a universally superior optimizer, but as a framework for determining when permutation search is useful, which representation and operator should be used, and when simpler alternatives are preferable.

Reinforcement learning with verifiable rewards (RLVR) has emerged as an effective approach for improving multimodal reasoning. However, most existing methods evaluate an entire response using a binary reward based only on final-answer correctness, thereby discarding the supervision available in intermediate reasoning steps. Process reward models offer finer-grained feedback, but they typically rely on separately trained verifiers, costly chain-of-thought annotations, or online judging by large language models (LLMs). In this work, we introduce StructReward, a compute-efficient framework that provides dense reinforcement signals through structured step-level reward alignment. StructReward represents each generated solution as a sequence of reasoning steps and aligns them with process-labeled reference steps using lightweight numerical, symbolic, and lexical matching rules. The aligned labels are aggregated into a dense process reward and combined with final-answer consistency and output-validity rewards through a gated Group Relative Policy Optimization (GRPO) objective. We further recycle policy rollouts into complementary supervision for response comparison and reflective self-correction, rather than discarding them after policy updates. Separately, we use a strong LLM to rewrite sampled correct trajectories into reflection-oriented training instances, further strengthening the policy's ability to evaluate and refine its reasoning. Since reward computation is performed online without an additional learned verifier or external LLM judge, StructReward substantially reduces the computational overhead of multimodal reinforcement learning. Experimental results show that structured process supervision and rollout recycling provide an efficient path toward self-improving multimodal reasoning.

Physics-informed neural networks applied to the level-set formulation of interface advection commonly augment the residual and initial-condition losses with an eikonal regulariser, penalising the deviation of $\|\nablaφ\|$ from unity. A previous two-dimensional study identified this weight as the dominant hyperparameter and found its optimum shifts by four orders of magnitude between rigid-body and deforming flows, but left open whether these principles transfer to three dimensions and whether single-seed results survive run-to-run variability. We answer both by repeating the weight selection across four 3D benchmarks (translating sphere, rotating sphere, slotted sphere, reversed vortex), sweeping six weights with three seeds at full training budget under a pre-registered selection rule. The ordering transfers: the selected weight tracks how far the exact solution departs from the signed-distance property, spanning four decades from $10^{-1}$ where it holds exactly to $10^{-5}$ where the interface is stretched. Values transfer only benchmark by benchmark; two of four carry over unchanged and two do not, so inheritance must be verified. The multi-seed protocol reveals that at small weights the seed-to-seed standard deviation equals the error itself, and the regulariser reduces it by more than an order of magnitude, buying reproducibility as well as accuracy. We benchmark against a fifth-order WENO solver on identical grids and error measures; the classical scheme is more accurate on all four problems, by two orders of magnitude on smooth rigid advection, with a margin that narrows with geometric difficulty and is smaller in volume conservation than in the field norm. Finally, we show that the relative $L_2$ error cannot certify the preservation of thin features, and report a feature-restricted measure that can.

Floods and landslides often co-occur, but their relationships with environmental controls vary spatially. This study develops a spatial heterogeneity-aware framework for flood-landslide susceptibility and relative-risk mapping in Kerala, India, and Nepal. It combines 15 km x 15 km grid cells with region-specific contextual zones and compares proximity-gated cross-zone training (S1) and ecology-gated zone-constrained training (S2). S1 permits geographically nearby models to be assigned across contextual boundaries, whereas S2 restricts model development and assignment to the same zone. Random Forest models for each hazard use strategy-specific predictor sets and are evaluated on spatially held-out test samples. Susceptibility surfaces are integrated with CRITIC-weighted exposure and vulnerability indices to produce hazard-specific and nine-class bivariate relative-risk maps. S1 achieved higher mean accuracy, precision, recall, F1-score, AUC-ROC, and PR-AUC for both hazards and regions. The largest difference occurred for Nepal flood susceptibility, where AUC-ROC increased from 0.728 under S2 to 0.886 under S1 and PR-AUC from 0.512 to 0.823. S2 produced lower Brier scores for both Nepal hazards and retained zone-specific differences in predictor selection, SHAP rankings, and response patterns, particularly in Kerala. Both strategies reproduced flood-prone lowland and landslide-prone upland patterns but differed in susceptibility and risk classes. Bivariate risk-map agreement was 0.521 in Kerala and 0.711 in Nepal, with allocation disagreement exceeding quantity disagreement in all S1-S2 comparisons. Susceptibility-to-risk correspondence remained below 0.350, showing that exposure and vulnerability changed priority locations. Overall, cross-zone learning strengthens regional discrimination, while zone-constrained learning preserves environmental differences, supporting their integration.

Brain magnetic resonance imaging (MRI) is central to neuroscience and clinical assessment, but models are commonly developed for individual diseases, populations or imaging protocols. Foundation models promise more general representations, yet they are usually pretrained once and can lose earlier capabilities when updated with new data. Here we show that Alcmaeon, a three-dimensional brain MRI foundation model pretrained without manual labels on more than 425,000 volumes and derived imaging maps, can be expanded sequentially across clinical domains. Alcmaeon combines volumetric encoding and latent diffusion generation with Graph-Blueprint Pruning (GBP), which protects network modules important to earlier domains while leaving the remaining capacity trainable. Across expansion from healthy ageing and neurodegeneration to developmental, psychiatric and tumour imaging, GBP showed less forgetting than sequential adaptation and elastic weight consolidation across voxel-level reconstruction measures, with its largest advantage after adaptation to tumour imaging. The blueprints provided an inspectable record of how model capacity was protected and reused. Representations from different model levels supported image synthesis, disease classification, survival modelling and postoperative prediction, although no single representation was optimal for every task. These findings provide a route towards brain MRI foundation models that can grow with emerging data while retaining earlier capabilities.

Biological neural systems achieve high efficiency and robustness through compartmentalized architectures. In contrast, modern artificial neural networks rely on globally entangled structures, which obscure decision logic and suffer from catastrophic forgetting. Here, we report a Decomposable Spiking Neural Network (D-SNN) that eliminates global synaptic entanglement by structurally isolating classification pathways into independent experts. Optimized via a bio-inspired push-pull loss function, the D-SNN achieves competitive accuracies on MNIST, Fashion-MNIST, and CIFAR-10/100 benchmarks. This modular approach matches the performance of fully dense networks while utilizing an order of magnitude fewer parameters. In addition, our networks operate with up to several orders of magnitude lower firing rates and fewer synaptic operations. Furthermore, physically severing connections between experts provides inherent protection against catastrophic forgetting during sequential learning. Crucially, these isolated pathways generate auditable neural signals, increasing decision transparency. This biomimetic, verifiable architecture establishes an efficient foundation for deploying deterministic neuromorphic intelligence in resource-constrained edge environments.

We present Ouroboros, a self-developing agent harness whose tools, prompts, context assembly, and core implementation improve through reviewed commits that become the runtime for later work. Core evolution proceeds in two modes. In recursive free evolution, improvement is itself a task, and completing one evolution cycle can schedule the next. In experience-driven core evolution, ordinary work and social interaction expose bugs, rough edges, and inefficient context construction that lead to reviewed structural changes. On Terminal-Bench 2.1, an Opus 5 run scores 86.74%, the best result reported on the benchmark. On OSWorld-Verified, an Opus 5 run reaches 90.69%, exceeding the best previously reported score. A five-rollout CL-Bench campaign achieves a normalized reward of 0.2301, setting a new state of the art. Hope is the longest-running publicly documented Ouroboros deployment. It is a 161-day living agent experiment in free evolution under governed human communication across seven surfaces. Human interaction surfaces faults and generates proposals, but the agent decides which changes to pursue. Because a self-developing agent may rewrite its own code and select new model APIs, operational safety becomes a primary design problem: guardrails must remain authoritative under evolutionary and public social pressure. Benchmark campaigns use frozen system snapshots, while Hope continues live evolution on a separate lineage.

We argue that learning visual representations without labels requires a training signal jointly complete across three non-overlapping objectives: semantic invariance across augmented views, patch-level spatial prediction, and representational non-degeneracy. We formalize these as the observation, prediction, and regularization principles and prove (i) that combining observation and prediction without regularization admits the constant encoder as a global minimizer under negative-free alignment; (ii) that the two objectives are gradient-complementary and structurally non-conflicting at the encoder output; and (iii) that the momentum encoder converges to the same fixed point as the online encoder and provides no collapse guarantee at convergence. Contrastive alignment provides only self-limiting collapse resistance, formalized via an explicit gradient-decay argument. Dropping prediction withholds the spatial training signal by construction; dropping observation forfeits cross-view semantic invariance by construction; at the scale we study, no pair substitutes for the third. Every major self-supervised method is a special case of a single unified energy decomposition. We pair every theoretical claim with a controlled experiment, including a patch-retrieval evaluation for the spatial consequence of prediction.

Modern vision systems must operate in "open-world" settings, where models must recognize known categories and detect unseen or anomalous content. Conventional semantic segmentation models operate under a "closed-world" assumption, often producing overconfident misclassifications on novel content. We address open-world semantic segmentation, the joint task of segmenting known classes while detecting and grouping novel or anomalous content without additional supervision, by extending a dual-decoder baseline with a third, complementary decoder within a unified encoder-decoder design. The first decoder performs closed-set segmentation using Gaussian prototypes for known categories. The second uses contrastive feature learning to isolate unknown regions in embedding space. The third, our key contribution, is a sensitivity decoder that captures fine-grained texture irregularities and activation instabilities indicative of semantic uncertainty, which neither semantic prototypes nor contrastive norms can reliably detect. The three decoders provide genuinely complementary signals: class-level OOD distance in logit space, global feature energy in embedding space, and local activation instability across encoder scales. Experiments on Cityscapes and BDD-Anomaly show that our method improves anomaly segmentation and novel-class discovery while maintaining competitive closed-set accuracy, with gains of +2.4% AUROC and a 2.5 pp. reduction in FPR@95TPR on BDD-Anomaly over the baseline.

Medical Visual Question Answering (VQA) requires aligning subtle visual evidence, including lesion texture, boundary sharpness, and diffuse density changes, with clinical language. Existing multimodal fusion approaches operating in the spatial domain may not fully exploit complementary frequency information present in visual and textual representations. We introduce a dual-branch frequency-domain fusion module that conditions spectral filtering on the input question, enabling adaptive selection of global low-frequency structure and fine-grained high-frequency detail before reconstructing the spatial representation for answer generation. To provide a richer spectrum for filtering, we extract complementary features from early texture-sensitive and final semantic layers of a frozen BiomedCLIP encoder and align both with the question representation using a symmetric InfoNCE objective prior to staged joint training with a BioBART decoder. We pretrain the proposed model on PMC-VQA and fine-tune it on the VQA-RAD and SLAKE benchmarks, demonstrating that frequency-aware multimodal fusion improves medical VQA performance while maintaining a lightweight and efficient architecture.

Agentic skills improve large language model (LLM) agents by encoding reusable procedures for complex tasks. However, manually authored skills often adapt poorly to long-horizon tasks and changing environments. To address the limitation, self-evolving skill systems have been developed to automatically construct and update skills from execution trajectories, shifting skill acquisition from external marketplaces to a trusted evolution pipeline. By replacing external skill acquisition with trusted internal construction, self-evolving skill systems reduce exposure to skill injection attacks that rely on direct skill manipulation. However, this skill evolution pipeline may introduce a new attack surface in which an attacker can indirectly steer skill evolution by inducing compromised trajectories through agent interactions. To demonstrate the threat, we propose Trajectory Backdoor Attack (TBA), a query-only attack that steers a trusted skill-evolution pipeline toward producing a backdoored skill. Specifically, we craft attacker-submitted queries to lead the agent to perform the target action and explicitly state the corresponding activation condition in the trajectory. We repeat the same condition-action pattern across diverse triggered tasks, while leaving clean queries unchanged, encouraging the evolver to consolidate the pattern as a reusable trigger-dependent rule into the evolved skill. Experiments on three benchmarks across two skill-evolution systems using four open- and closed-source backbone models demonstrate that TBA reliably implants conditional backdoors while preserving clean-task utility, matching or even surpassing direct skill injection. The results reveal a critical vulnerability in trajectory-driven skill evolution.