The Inference Report

Recent advances in inference-time scaling have significantly improved the reasoning performance of large language models (LLMs). However, these methods typically rely on repeated generation or external verification. To address this limitation, we introduce CritICL, a novel inference-time framework that improves reasoning while maintaining high efficiency. Our key insight is that LLM failure modes exhibit structured patterns across model scales within the same family. Instead of treating failures as undesirable outputs, CritICL leverages them as a source of guidance. Specifically, we utilize failure modes derived from weaker models and incorporate them into inference through critique-based in-context examples. We propose two variants: CritICL-dynamic, which adaptively predicts input-specific failure modes and retrieves critiques, and CritICL-static, which uses a global failure mode profile to provide stable guidance. Experimental results show that CritICL consistently outperforms standard in-context learning and achieves performance competitive with or superior to test-time scaling methods, while requiring significantly fewer generations and lower token cost. Code available at: https://github.com/umwyf/CRITICL

Agent skills package specialized knowledge and workflows into reusable resources that extend AI agent capabilities. Recent work automatically discovers such skills from agent experience, which enables agents to progressively adapt through interaction. However, the insights that guide skill development typically remain scattered across optimization histories, limiting their systematic reuse across iterations. We introduce WikiSkill, a framework that co-evolves agent skills with a persistent knowledge base (wiki). At a high level, WikiSkill separates raw execution experience, accumulated knowledge, and executable skills, while continuously consolidating experience into the wiki, which subsequent skill updates can build on. Across diverse benchmarks and models, WikiSkill consistently outperforms state-of-the-art skill-evolution methods and improves over no-skill baselines in most model-benchmark settings. We find that skill evolution complements model scaling: larger models generally benefit more from evolved skills, while smaller models with skills can outperform substantially larger models without them. We also find that evolved skills transfer effectively across models and model families, and skills evolved by other models can outperform self-evolved skills. Finally, our ablation studies confirm that persistent knowledge accumulation in the wiki is critical for effective skill evolution. These results demonstrate the benefits of systematically accumulating and refining agent experience for developing reusable and transferable skills.

Empirical comparisons between systems are a standard form of evidence in computer science research, but few are checked for statistical validity: most are never framed as statistical tests at all. Existing multiple-comparison procedures could control the resulting error, but need inputs (what an analysis examined, and how its observations are arranged) that are not recoverable from a list of p-values. We introduce Tacet, a language in which an analysis declares what it generated, states what it expects to find, and is refused any claim it cannot afford or cannot properly test. Its core calculus T pairs a free estimation sublanguage, carrying a reported footprint and a purity bit that records whether any outcome was consulted in building a value, with a priced claim sublanguage, carrying a wealth transformer, connected only by a mechanism that prices a comparison. A sample selected by reading outcomes sets the purity bit and is recorded as having examined everything it read, permanently, so it can never be granted a one-sided or confirmatory price, without the system ever asking whether the analyst intended to cherry-pick. Whether a comparison is paired or clustered is computed statically from the artifact schema, from declared functional dependencies between key fields alone and before any data is read, and a mechanism that assumes that structure away is refused rather than priced. Because the wealth transformer is antitone in the realized p-value, affordability can be checked before the analysis runs too, turning pre-registration into a typing rule. We prove the metatheory machine-checked in Lean 4 with no admitted gaps, and demonstrate the approach on a reference implementation and two case studies on published artifacts, the SWE-bench Verified leaderboard and BIG-Bench Hard.

To improve large language models' ability to resolve real-world software issues, prior work has focused on constructing large-scale agent trajectory datasets and performing supervised fine-tuning (SFT) on successful trajectories. However, task success does not guarantee high-quality supervision: successful trajectories may still contain ineffective, redundant, or risky steps. Directly using such trajectories for SFT can introduce noisy supervision and encourage models to imitate undesirable problem-solving behaviors. Therefore, we propose SWE-Prime, a multi-granularity, two-stage SFT data selection method that progressively filters training data at the trajectory and segment levels. Specifically, the first stage performs trajectory-level screening based on process quality, result quality, and data representativeness, selecting a high-quality and representative subset of successful trajectories. The second stage performs segment-level selection by grouping consecutive steps into semantic segments and assessing each segment based on its contribution to the final solution, learnability, and potential risks. During SFT, all segments remain in the sequence to preserve context, while only selected segments contribute to the loss computation. Experiments on SWE-Bench Pro and SWE-Bench Verified show that training on the 10% trajectory subset selected by SWE-Prime outperforms training on the full resolved dataset, yielding relative performance gains of up to 12.2% and 24.2%, respectively.

Recent prominent post-training methods, such as Reinforcement Learning (RL) and On-Policy Self-Distillation (OPSD), have driven rapid progress in mathematical reasoning for large language models, yet their reliance on ground-truth labels precludes test-time training (TTT). Replacing ground truth with majority-vote pseudo-labels is a natural alternative, yet it is fragile: an incorrect vote corrupts the teacher and misleads every token. We observe that this failure mode is asymmetric: rollouts that disagree with the pseudo-label are typically wrong regardless of whether the vote itself is correct. Building on this observation, we propose Test-Time Policy Optimization (TTPO), an asymmetric objective that distills agreeing rollouts via OPSD and penalizes disagreeing rollouts with Grouped RL. Token-level selection further refines both branches: distillation down-weights already-converged positions, while RL penalizes only confident errors. Both updates remain well-grounded even under frequent pseudo-label errors, and majority-vote routing yields tighter self-supervision as the model improves. Without any labels, TTPO matches label-supervised OPSD on five competition-level benchmarks, raises Qwen3-1.7B from 38.0% to 45.2% in TTT, yields +25.2% to +36.4% without thinking, and shows strong cross-task generalization.

In real-world software development, code review typically involves iterative interactions between developers and reviewers to improve software quality, making the process costly and time-consuming. Although recent work explores large language models (LLMs) for automated code review, most approaches oversimplify code review into a single-round, static decision task, which fails to capture the multi-round interactive nature and the complex problem-solving processes inherent in realistic review scenarios. To bridge this gap, we introduce MCR-Bench, the first defect state-aware benchmark designed for realistic multi-round code review. MCR-Bench covers five commonly-used programming languages and consists of 2,269 real-world multi-round code review tasks, each of which is annotated with fine-grained defect information and cross-round state labels. Each task in MCR-Bench is equipped with fine-grained defect metadata (e.g., description, type, severity) alongside dynamic state annotations, capturing the complete evolutionary trajectory of a defect throughout the multi-round process. We obtain several findings through extensive experiments on MCR-Bench with mainstream LLMs. (1) Limited overall capability: experiments reveal that mainstream LLMs exhibit limited overall performance in defect detection and defect lifecycle state tracking, with performance degrading significantly as the number of interaction rounds increases; (2) Defect-sensitive performance: LLMs' performance varies substantially across different defect types and severity levels, with semantically complex or low-salience defects being significantly more likely to be missed; (3) Underlying Failure Mechanisms: our in-depth error analysis dissects the distinct drivers of false positives and false negatives, revealing critical weaknesses such as cross-round temporal misalignment and inadequate long-range memory.

LLM-based agents are increasingly deployed in product-level execution harnesses, where jailbreaks can trigger harmful tool use and persistent state changes, creating greater risks than unsafe text generation alone. Existing automatic red-teaming methods often rely on fixed attacks, while recent agentic attackers coordinate multiple jailbreak tools and show stronger potential through trajectory-based retrieval. However, such retrieval can reuse misleading experiences due to retrieval bias and unclear tool credit, and full trajectories add context overhead while reducing interpretability. We propose RedEvoAgent, a black-box red-teaming agent that distills cross-case attack trajectories into a concise, human-readable attack skill. The attack skill adaptively evolves through tool-effectiveness profiling and Deciding-Tool Attribution for skill updates, and a validation ratchet that retains only updates improving validation performance. Experiments on multiple benchmarks, target models, and target execution harnesses show that RedEvoAgent outperforms fixed and agentic baselines, improves tool efficiency, and transfers across attacker models and target execution harnesses.

Chemical reactions are fundamentally transformations in electron space, yet most machine learning approaches model them either through \textit{de novo} generation of product molecules or through heuristic graph edits that operate directly on molecular topology. We introduce MAELLE (\textbf{M}ech\textbf{A}nistic \textbf{E}dit f\textbf{L}ow-matching on e\textbf{L}ectron r\textbf{E}arrangements), which instead models reactions as discrete flow matching over electron occupation vectors. Concretely, we formulate the reactant-to-product mapping as a Continuous-time Markov Chain (CTMC) over the graph-structured integer-valued electron occupation space defined on all bonding, non-bonding, and hydrogen sites. To construct the intermediate edit trajectories, we generalize the discrete flow matching mixture path to discrete electron rearrangements using Optimal Transport, yielding a sequence of mechanistically interpretable edit moves without requiring elementary step annotations. MAELLE achieves competitive performance on the USPTO-480K benchmark compared with leading reaction prediction models. Beyond in-distribution accuracy, we evaluate robustness across two out-of-distribution settings - structural complexity and reaction type - and find that MAELLE maintains strong performance where existing methods degrade. Finally, because the learned flow operates over the full electron redistribution, MAELLE naturally recovers mechanistic trajectories that align with known chemistry and can predict side products of a reaction.

Transduced language models (TLMs) compose a pretrained \emph{source} language model with a functional finite-state transducer to induce a language model over \emph{target} strings. Computing the probability of a target prefix under a TLM amounts to summing the source-model probabilities of all source strings that the transducer maps to target strings beginning with that prefix. This set can be exponentially large or infinite. Prior work uses a computational shortcut based on source prefix probabilities, then approximates the resulting sum with threshold-pruned beam summing. This produces a lower bound with unknown error. Instead, we resample source prefixes without replacement and reweight each selected prefix by the inverse of its inclusion probability. We show that applying this correction recursively gives an unbiased estimator of the target prefix probability and lets us estimate the mass lost by threshold pruning. Our beam-summing algorithm extends the retained source prefixes and samples which prefixes to keep, reducing their number as more probability mass is added to the running estimate. This can save computation and guarantees that the run halts with probability one. We evaluate the method on encyclopedic text and DNA against sequential Monte Carlo baselines that resample with replacement. It achieves a better compute--variance tradeoff on text and lower error at the same maximum number of particles on DNA. On a DNA-to-amino-acid transduction, it reduces runtime by several orders of magnitude relative to threshold-pruned beam summing and makes estimating prefix probabilities for long target strings feasible. Replacing threshold pruning with unbiased sampling in a published reading-time analysis substantially lowers the estimated corpus surprisal but leaves the published conclusions unchanged.

Large language model (LLM) agents in governed organizations must let the persona (instructions, tone, self-presentation) evolve freely, while keeping execution (stateful, audited work) traceable. A single trust domain does not satisfy both cheaply. We present Persona-Execution Separation (PES): persona and execution reside in different trust domains, connected by a governed contract bridge. The persona is singly-homed and may drift; execution is faceless and audited. Status summaries may return; data bodies remain in the restrictive domain except a graded data-loss-prevention (DLP) exception; identity stays continuous. An approval matrix, DLP, and audit enforce the crossing. PES follows from three goals---free drift, execution traceability, and decoupling. Under LLM representational indistinguishability, any single-domain mechanism that meets all three must re-introduce typed change objects, an external gate, and a stable audit anchor: PES rebuilt at higher coupling cost. A development/pilot case in a regulated digital-employee platform records five decisions over one month, each with a rejected alternative. A mechanism check on the shipped implementation found no execution-side re-validation under persona perturbation (five model configurations) and no persona fingerprint on hard-asserted fields. A probe of a recovered pre-separation build found the governed execution path decoupled from the persona by omission, not by construction; a later wiring change could reverse that isolation, which PES makes an audited architectural rule. The pattern applies when multi-user deployment, execution audit, and expected persona churn hold jointly.

Static scanners are increasingly used to identify executable or otherwise unsafe content in machine- learning artifacts, yet conventional evaluation metrics characterize only cases where a scanner yields a usable security judgment. We evaluate ModelScan, ModelAudit, and Fickling using a controlled, artifact-backed benchmark on a synthetic corpus of 170 Pickle and PyTorch focused artifacts across 145 specimen families, 135 of which have binary security ground truth and 10 of which are intentionally malformed without labels. We explicitly distinguish non-N/A coverage, analysis completion, definitive security decisions, non-security findings, and unsupported outcomes. On labeled families, ModelAudit produced definitive security decisions for all 135 families (100%), Fickling for 110 (81.5%), and ModelScan for 67 (49.6%). Conditional on making a definitive judgment, ModelScan achieved 100% precision, recall, and F1. Fickling identified no unique true- positive families beyond those found by the combination of ModelAudit and ModelScan. Furthermore, for the 48 malicious families where ModelScan failed to complete its analysis, both ModelAudit and Fickling generated detections consistent with ground truth. These findings underscore the need to separate judgment accuracy from judgment availability, as well as incremental detection coverage from tool-level redundancy.

Currently used sepsis severity indices rely on fixed variables and weights established decades ago, which are coarsely discretized and calibrated to a cohort that no longer reflects contemporary critical care. No alternative learned directly from patient trajectories is in routine use. We conducted a retrospective two-cohort study on a total of 29,116 and 7,691 adult patients meeting Sepsis-3 criteria from two hospital systems in Massachusetts and Georgie, respectively. We developed a sepsis index using 43 routinely charted variables over a 72-hour treatment window. Unlike previous studies, we use mortality as a treatment-level ranking signal rather than a per-state target, allowing credit to be redistributed non-uniformly across timesteps. Evaluation was done on a permanent 20% test holdout, using clinical vignettes and Spearman correlation. Uncertainty intervals were obtained by bootstrap resampling of whole patients. Under this ranking scheme, non-survivors scored 1.19-1.64 points higher than survivors on a 0-10 scale within all strata of baseline SOFA-2, with similar results stratifying within lactate, mean arterial pressure (MAP), and creatinine. Within-patient change in the index correlated with change in lactate (Spearman rho = 0.39; n = 1,854). Similar, weaker correlations were found for MAP and creatinine. On a cohort level, cross-institutional agreement measured by Spearman correlation between models trained on different sites, were 70-77% of same-site correlation. External within-patient correlations were 0.54 and 0.59 against ceilings of 0.92 and 0.90. Our index also correlated with established indices, while null controls stayed near zero. Our index demonstrated hourly prognostic information that meaningfully separates patient outcomes and is consistent with clinical expectation, indicating potential as a decision support tool complementing clinical judgement.

Reinforcement Learning with Verifiable Rewards (RLVR) significantly improves LLM reasoning but often causes a drop in policy entropy, leading to narrowed reasoning coverage and degraded pass@$k$ for large $k$. While existing methods mitigate this entropy collapse through algorithmic regularizations, cross-model non-parametric perturbation is also neglected. In this work, we propose a simple yet effective approach to preserve the generative diversity of LLMs during RLVR. Instead of relying solely on internal exploration, we force the target model to generate answers based on partial reasoning trajectories generated by a smaller, weaker language models. These unfamiliar prefixes effectively disrupt over-confidence and encourage the exploration of distinct reasoning paths. We empirically study the potential of outer prefixes, revealing the mechanism of the impact of distributional discrepancy to the exploration dynamics in RLVR training. Experiments across multiple mathematical benchmarks show that our method consistently outperforms vanilla RLVR. Notably, the performance gain becomes increasingly pronounced as $k$ scales up, demonstrating a substantial expansion of reasoning coverage. Furthermore, our approach efficiently mitigates entropy collapse without requiring additional SFT, intricate reward designs, or complex prompting.

Friend recommendation is inherently graph-structured: the relevance of a potential connection depends on multi-hop social context rather than user attributes alone. However, deploying message-passing GNNs on a production-scale social graph with hundreds of millions of users and tens of billions of edges requires addressing numerous modeling and systems challenges. We present a scalable end-to-end GNN ranking system for production social graphs, focusing on two design choices that are critical in this setting: multi-hash ID embeddings and temporal neighbor sampling. Multi-hash embeddings are common for high-cardinality features, but industrial GNN systems typically either ignore trainable IDs or accept full embedding tables, exceeding 200 GB for our graph. We integrate multi-hash as the primary node representation, reducing the ID-embedding table size by more than 98 percent while preserving ranking quality. Temporal neighbor sampling is well understood in principle, but existing implementations scan full adjacency lists, which is a non-starter for users with tens of thousands of friends. We implement timestamp-sorted CSR storage with binary search, reducing the per-node temporal sampling cost from $O(deg(v) + k)$ to $O(\log(deg(v)) + k)$. Beyond these components, we show that this combination scales and yields measurable production impact. On a graph with 194M users and 28B edges, offline ablations isolate each design choice's contribution. In an online A/B test, our system increases friend additions from recommendations by 16 percent and unique friend adders by 11.5 percent over a strong production baseline. We release our framework for distributed training and inference on large temporal graphs.

Reinforcement learning with verifiable rewards (RLVR) improves specific capabilities of large language models, but covering multiple capabilities often involves training separate domain experts and subsequently consolidating them. We organize three fusion paradigms by the artefacts they reuse: Merge combines expert task vectors, Mix RL pools their datasets, and multi-teacher on-policy distillation (MOPD) uses both. Because they have largely been studied in isolation, how they compare and how to choose among them remain unclear. We compare all three using shared experts and data across model scales and a multi-domain benchmark suite. Although their average performance differs by at most 1.4 points, the gap reaches 8.6 points on a single benchmark, with domain-level variation tracking cross-domain relations visible in task-vector geometry. Training dynamics expose distinct constraints: Mix RL depends on domain mixture proportions, MOPD remains bounded by its teachers, and Merge compresses all expert updates into one. All three improve single-sample accuracy without measurable gains in solution coverage or losses in held-out capabilities. These results yield a practical guideline: use Merge when experts already exist and cheap fusion is paramount; Mix RL when training a unified model without experts, with domain proportions adjusted for cross-domain transfer; and MOPD when preserving domain-specific gains matters more than surpassing teachers or minimizing end-to-end cost.

State-of-the-art action-conditioned video models are typically restricted to a single robot embodiment, preventing them from leveraging the vast corpus of heterogeneous video data that contains rich signals for learning generalizable physics. To bridge this gap, we introduce CLAP, a framework for cross-embodiment action-conditioned video generation capable of being trained on diverse, internet-scale videos across human and robotic agents. CLAP is grounded in the insight that universal physical laws govern spatiotemporal dynamics regardless of the actor. However, cross-embodiment learning is non-trivial because action representations vary sharply across robot platforms and are typically absent in human videos. CLAP addresses this fundamental challenge through the following core contributions. First, CLAP reconciles disparate action spaces using end-effector poses, language instructions, and latent actions. Second, to resolve their individual limitations, CLAP introduces a curriculum-based cross-embodiment learning recipe that first learns foundational physical priors across unlabeled video data using latent actions and subsequently grounds them in end-effector action spaces for zero-shot deployment to real-world tasks. Crucially, CLAP approaches or surpasses state-of-the-art single-embodiment video models in challenging environments like DROID. These performance advantages compound via few-shot adaptation to establish a novel paradigm for training single-embodiment video world models. Ultimately, CLAP delivers the most comprehensive suite of action-conditioned video world models to date - spanning diverse action-conditioning spaces (end-effector, language, and latent) and robot morphologies (including cross-embodiment, DROID, Bridge, bimanual YAM robots, and G1 humanoids). We open-source all code and models. Project Website at https://omni-clap.github.io .

How do large language models (LLMs) organize moral knowledge? Models detect moral content broadly, but detection is a low bar. We ask whether they go further, distinguishing moral foundations from one another and organizing the relationships between them geometrically. We train six independent linear probes on open-weight language models, one per Moral Foundations Theory (MFT) category (care/harm, fair/cheat, lib/oppress, loy/betray, auth/subv, sanc/degrade), and examine how the resulting directions relate to each other in representation space. We find the directions neither collapse into a single moral detector nor isolate from one another. Rather, they span a near-maximal number of independent dimensions while sharing a positive common component. The shared component is the signature of integration, and it is moral-specific relative to a matched non-moral concept battery built identically (mean pairwise cosine 0.26 vs. 0.013). The geometry is consistent across architectures and scale and reaches its integration regime early in pre-training, well before probe accuracy saturates. The structure the model discovers shows no evidence of the individualizing/binding distinction predicted by Moral Foundations Theory (an underpowered test: only 20 candidate partitions exist) but rather reflects corpus statistics. Extending to moral dilemmas, each dilemma direction partially composes from its component foundations, at 2.7x a mismatched-pair baseline, while the majority of its variance encodes conflict-specific structure. The model represents moral tension itself, not a pre-resolved judgment.

Clinical language models can achieve strong in-hospital accuracy yet fail under deployment shifts because they exploit note-specific artifacts (e.g., templates, separators, boilerplate) that do not reflect patient state. We propose CAST (Concept-guided Artifact Suppression Tuning), an SAE-based framework for auditable clinical text classification. CAST uses Sparse Autoencoders to expose sparse, human-auditable features from intermediate Transformer activations, labels SAE latents with an LLM-assisted interpretation pipeline and ICD-10 retrieval constraints, suppresses verified artifact latents via residual subtraction during fine-tuning, and provides post-hoc per-concept attributions for auditing model decisions. On MIMIC-IV discharge-note mortality prediction, CAST improves over its corresponding fine-tuned encoder baselines and remains competitive with strong LLM baselines, while producing a feature-level audit trail of the clinical concepts that support each prediction and the artifact concepts suppressed during training.

Video carries the temporal structure of the physical world, yet learning representations from it has remained computationally expensive: prevailing self-supervised methods either prevent representation collapse through architectural asymmetries, coupling an exponential-moving-average target encoder, a stop-gradient, and a capacity-limited predictor, or circumvent it by reconstructing masked content in pixel space. We introduce LeVJEPA, the first video encoder trained under LeJEPA's collapse-free objective, which dispenses with both. A single encoder is trained with an invariance loss over global and local views of a clip, regularized by SIGReg, which excludes collapse with a provable guarantee. The architecture reduces to an encoder and a projector, and the objective to a single hyperparameter. This formulation admits two properties. First, the cost of pretraining is governed by the number of tokens the encoder observes; uniform random token dropping renders this number small while simultaneously improving downstream accuracy. At matched epochs on identical data, LeVJEPA matches or surpasses V-JEPA 2 across ViT-S/B/L at 5.6 to 20.8x less pretraining compute, and at matched total FLOPs it exceeds the strongest video baseline by 7.6 points on ImageNet-1K while remaining competitive on motion-centric benchmarks. Second, since no asymmetry between branches is required, the encoder can be trained with block-causal attention at no measurable accuracy cost: temporal ordering becomes a property of the encoder itself. Against a compute-matched DINOv2 trained on frames of the same videos, LeVJEPA approaches the image-pretrained encoder on appearance-centric evaluation while nearly doubling its motion-centric accuracy. These results indicate that, once its computational overhead is removed, video becomes a viable and in several respects preferable substrate for general-purpose visual pretraining.

Retrieved scientific literature can serve as inspiration for both human and AI scientists. Inspiration can take different forms: prior work may directly suggest how to address a problem, or surface directions at different levels of abstraction - zooming out to a more general view or zooming in to a concrete realization. We introduce RATIO (Retrieval Across Typed Ideation Operations), a large-scale benchmark in which relevance is defined by three operations which we name ideation moves: Address retrieves potential approaches for stated problems, Broaden retrieves more general formulations, and Specify retrieves concrete instantiations. RATIO is constructed from millions of full-text scientific papers across CS literature via a general recipe that extends discourse-marker distant supervision - previously used only for classification - to corpus-scale retrieval, combined with extensive LLM and human vetting. Experiments show that operation-specific fine-tuning substantially boosts retrievers but leaves much room for further improvements. RATIO provides a scalable training and evaluation framework for retrieval components that support literature-grounded ideation, opening up new research avenues on scientific inspiration retrieval.

Camera-derived remote photoplethysmography (rPPG) is commonly validated through endpoint accuracy, but endpoint performance does not establish whether other physiological properties of source contact photoplethysmography (PPG) remain preserved recording by recording. We evaluated property-specific PPG-to-rPPG recoverability on 655 recordings from the Multi-Domain Mobile Video Physiology Dataset using CHROM as a fixed camera-rPPG observation pathway. The pathway reproduced the published CHROM correlation regime, with heart-rate MAE of 15.26 bpm and Pearson correlation of 0.0801. Matched-versus-shuffled validation revealed modest recording-specific autocorrelation correspondence, while spectral and recurrence-rate measures showed little matched discrimination. Maximal Lyapunov exponents showed essentially no recording-specific PPG-to-rPPG correspondence, with correlation of 0.0231 and permutation p-value of 0.5584, despite population-level overlap. Endpoint discrepancy exhibited Fitzpatrick-associated heterogeneity after adjustment for lighting and motion, including a Fitzpatrick VI versus III contrast of 9.32 bpm, while dynamical discrepancy showed no corresponding gradient. Aggregate RGB signal-to-noise ratio did not materially account for the endpoint contrast. In subject-held-out analysis, adding motion and lighting consistently reduced MAE across linear, ridge, and random-forest learners relative to rPPG-HR-only calibration, with reductions up to 13.32 percent. These findings show that recoverability is property-specific: physiological properties differ in recording-specific preservation and dependence on observation conditions, and population-level plausibility does not establish preservation of individual recordings.

LLMs are increasingly able to answer complex questions about enterprise-scale document collections. But evaluation is hard: companies don't want to share internal communications, and synthetic datasets have been overly simple. We present CorporateBench (CB), a human-validated multi-task Q&A benchmark whose scale approaches the conditions LLMs encounter in corporate communication networks, with evaluation corpora surpassing 230,000 documents. CB evaluates LLMs across two dimensions (information extraction and knowledge base querying) through four synthetically generated firms ranging from 12 to 10,000 employees. Each corpus is sampled from a temporally evolving knowledge base describing a consistent world, guaranteeing cross-document logical consistency even across hundreds of thousands of documents. We evaluate five LLMs on CB, revealing increasingly poor performance as input size approaches realistic scales. CB provides LLM developers a metric for corporate communication reasoning, filling a crucial gap in the benchmarking ecosystem.

Generative AI is transforming how people access information, challenging traditional advertising mechanisms built around predefined slots. Towards generation-native advertising, we propose the Latent Advertiser Mixture Auction (LAMA), a token-level advertising mechanism that embeds advertiser influence directly into the generation process. Advertisers report local continuation values that induce advertiser-specific next-token policies, from which the platform decodes through a latent mixture while updating an allocation posterior. We show that LAMA satisfies Markov DSIC and IR, and achieves near-optimal KL-regularized welfare. We further develop a learning-based implementation that reconstructs the required reports online from learned local advantages and root values. Proof-of-concept experiments on real-world commercial-search query splits show that LAMA improves platform welfare and revenue while maintaining user-facing response quality, providing initial evidence for the feasibility of generation-native advertising.

AI-generated text detection is commonly framed as a binary document-level judgment about whether a text is human-written or machine-generated. This framing breaks down for mixed-origin writing, where content origin and expression origin may differ. We cast mixed-origin detection as dimension-to-composition source attribution, inferring content origin and expression origin before composing them into four collaboration types. We propose Dimension-to-Composition Routing (D2C-Routing), which routes content-side and expression-side evidence to supervised dimension heads before a learned gated composition layer predicts the final label. On MixD2C, a reconstructed split derived from the HART mixed-origin benchmark, our disclosed D2C-Routing-based detector system reaches 0.8603 four-way Avg TPR@1%FPR, 6.5 points above the same-split RACE-local rerun. Core ablations support the routing design, while error analysis shows that distinguishing AI-content/human-expression from fully AI-generated text remains the hardest boundary. Code is available at https://github.com/bystander563/d2c-routing-artifact.

We establish a sharp phase transition for fitting random vectors by an ellipsoid. The random vectors have independent subgaussian coordinates with mean zero, variance one, and a common fourth moment, and the number of vectors is proportional to the square of the dimension. We identify an explicit satisfiability threshold such that, with high probability, a positive definite ellipsoid passes through every data point below the threshold, whereas no positive semidefinite fit exists above it. We also determine the optimal squared fitting error throughout the unsatisfiable regime. In particular, the threshold depends on the coordinate distributions only through their common fourth moment, revealing a fourth moment universality phenomenon. For standard Gaussian data the threshold is $1/4$, resolving the ellipsoid fitting conjecture.

Language model pretraining has become almost synonymous with prohibitive cost, placing it out of reach for much of the academic and open-source communities. Although strong open-source efforts already exist, including open-weight models and open-source training recipes, a cost-efficient, hardware-accessible, and open-source pretraining recipe has long been missing. Even at a small scale, training Llama-3.2-3B costs over \$1.5M, and reproducing SmolLM3-3B needs over \$700K. In this report, we present an open pretraining recipe designed to lower this barrier. Using this recipe, we train a collection of Puro-2B models from scratch on up to 1.4 trillion tokens with FP8 precision on consumer-grade RTX 5090 GPUs. The models in the collection differ in token budgets and selected recipe variants. Our best model is trained at a compute cost of less than \$6.9K and approaches Qwen2.5-1.5B performance under our evaluation protocol. This cost efficiency is enabled by a combination of approaches, including hardware selection, low-precision training, hyperball optimization, curriculum model averaging, and the data recipe. Beyond the recipe itself, we provide two additional results. First, across the Puro-2B collection, we derive a Puro Cost Scaling Law that relates training cost to average model performance; the fitted law suggests that about \$4.4K, less than \$5,090, is sufficient to reach the performance of Qwen2-1.5B. Second, as an end-to-end case study, we examine how pretraining data curricula shape downstream performance after post-training. Such controlled studies are enabled by having access to the full pretraining pipeline rather than model weights alone. We release the full training recipe for Puro-2B, including data, code, and model weights under Apache 2.0 at https://huggingface.co/collections/thu-pacman/puro-2b.

Joint-Embedding Predictive Architectures (JEPAs) for world modeling typically employ fixed-size Vision Transformer encoders that are over-provisioned for simple tasks and under-provisioned for complex ones, with significant redundancy across attention heads. We propose Successive Capacity Growth (SCG), a method that starts from a minimal encoder (1 head, 2 layers, 283K parameters) and grows incrementally in width (adding attention heads for low-level semantic capacity) or depth (adding transformer blocks for higher-order semantic abstraction), driven by a task-agnostic test-and-verify mechanism that exploits function-preserving expansion to safely trial architectural changes and roll back if they do not improve prediction loss. The Sketched Isotropic Gaussian Regularizer (SIGReg) ensures that all learned semantic dimensions remain statistically independent and aligned with the predictive objective, preventing collapse even as the architecture grows. On a 60-dimensional multi-object dynamics task, SCG naturally triggers depth expansion, improving prediction loss by 20.3% over the fixed small baseline with 56 times greater parameter efficiency than scaling to the fixed large model; on a 2D navigation task, a single width expansion yields even an 23% improvement over the fixed large model. Across all three tested environments of increasing complexity, the adaptive encoder matches or exceeds the fixed small baseline, with zero false-positive expansions and bit-exact function preservation (ratio = 1.0, absolute difference = 0.0). The take-away is that JEPA world model encoders need not be pre-allocated at maximum capacity - they can grow successively as the task demands, achieving significant compute and data efficiency while maintaining representation quality.

Signals are short textual or visual snippets displayed on the eBay View-Item (VI) page, providing additional, contextual information for users about the viewed item. The aim of displaying these signals is to facilitate intelligent purchase and to incentivize engagement. In this paper, we present a 2 stage xgboost based model that optimally populates the VI page with signals. This approach has shown a 0.08% lift in overall GMB (Gross Merchandise Bought) and 0.58% increase in Parts and Accessories GMB, primarily due to increase in conversion of high average price items in online experimentation.

Physics-integrated 3D Gaussian representations now allow reconstructed deformable objects to be simulated and rendered under explicit material models. Existing pipelines, however, assume that material parameters are known or manually specified, limiting their applicability when these parameters must be inferred from observed object dynamics. We propose KnockGS, an interaction-response PhysicalGS framework that estimates the elasticity and density scales of a 3D Gaussian object from its dynamics under a known applied force. Rather than treating physical simulation only as a forward process, we turn the force-induced response into a calibration signal: temporal response features are xtracted from the observed dynamics, the two material scales are estimated from those features, and the estimate is then frozen and written back into the same simulator so that it can be tested on an interaction it was never fitted to.We evaluate the framework on both parameter recovery and response-level fidelity. The estimated scales are compared against hidden ground truth, and the re-simulated object is measured against the target using 3D particle trajectories, response-curve statistics, and rendered-frame quality. Across five held-out material targets, our method recovers the scales substantially more accurately than response retrieval, global regression, or a fixed default material, and the frozen estimate remains predictive under interactions that differ in direction and in magnitude. Interaction response therefore carries enough information to calibrate material scales in physically grounded 3D Gaussian representations.Our study is a first step toward interactive PhysicalGS systems that calibrate a Gaussian asset whose rendered appearance and simulated response are consistent.

We study how sophistication in generative AI (genAI) use varies among the back-office workforce of a large firm. Using proprietary data, we observe 713,564 employee prompts and their corresponding large language model responses from nearly 4,000 back-office employees across 15 functional areas over eight months in 2025. We document three main findings. First, senior employees exhibit more sophisticated genAI use, consistent with domain expertise complementing genAI capabilities. Second, sophistication varies considerably across functions and is highest in Strategy, Digital Innovation, and Project Management, three groups that share a focus on firmwide strategic initiatives and organizational change. Third, we observe neither improvements in sophistication over time nor lasting improvements following formal AI training, suggesting that sophisticated use can be difficult to change. Together, our study provides measures of and insights into sophisticated genAI use that managers can use to improve outcomes and that researchers can use in future research.

Speaker anonymization suppresses speaker-identifying attributes from speech while preserving linguistic content and quality. We propose repurposing XTTSv2, a multilingual voice cloning model trained on 27k hours of speech, for speaker anonymization without retraining. Our key insight is that XTTSv2's voice cloning capabilities preserve prosodic structure independently of speaker identity, enabling voice conversion by conditioning on a pseudo-speaker. We introduce an iterative refinement strategy that balances privacy and utility by maximizing a harmonic mean of speaker dissimilarity and intelligibility. Evaluated on seven European languages across CommonVoice and Multilingual LibriSpeech, our system achieves near-optimal privacy (EER $\approx$ 0.49), competitive intelligibility, and substantially better speech quality than dedicated anonymization baselines, while requiring no language-specific training. We release the code here: https://github.com/rm00cr/coqui-tts.

Influential public discourse shapes public beliefs and can also mislead, not only through what is stated, but also through how information is framed, omitted, contextualised, and communicated. Yet less research has focused on how such misleadingness arises and shapes the interpretations formed by readers. To address this gap, we introduce Reader-Centric Misleadingness Understanding (RCMN), a framework that operationalises misleadingness through five dimensions: misleading mechanism, likely reader interpretation, evidence-warranted interpretation, emotional arousal, and communicative intent. Based on this framework, we construct an evidence-grounded dataset of influential public discourse. Empirical findings show that misleadingness is diverse and extends well beyond fabrication, with unsupported inference, exaggeration, and omission among the prevalent mechanisms, and is frequently associated with heightened emotional arousal and distortive communicative intent. Moreover, we investigate whether lightweight claim-and-context representations retain sufficient cues for understanding reader-centric misleadingness without access to richer contextual, evidential, and multimodal information. Evaluation across five recent generative foundation models shows that reader-level interpretations can often be recovered from such limited representations, whereas identifying how misleadingness is produced remains considerably more challenging. These findings highlight the potential of lightweight representations for scalable misleadingness analysis, while reliable understanding of misleading mechanisms continues to require richer contextual and evidential grounding.

Evolution Strategies (ES) have recently emerged as a memory-efficient post-training paradigm for LLM reasoning. However, the optimization behavior of ES remains understudied, making it hard to define its advantage scope compared to mainstream post-training paradigms (e.g., Group Relative Policy Optimization (GRPO)). By systematically investigating ES dynamics and mechanisms, this paper first identifies a performance advantage of ES over GRPO, theoretically and empirically showing that ES can lead to broader reasoning coverage, thereby better exploiting the reasoning capabilities of pretrained LLMs. Theoretically, we show that verifier-projected Jensen-Shannon diversity across the ES population is helpful to higher Pass@K performances. Empirically, unlike GRPO, which exhibits entropy collapse, ES improves Pass@1 while attaining higher Pass@K than GRPO. We further develop a sequential GRPO-ES training strategy that combines GRPO's strength in Pass@1 with ES's gains in Pass@K. Second, we find that despite substantial whole-model parameter drift, the task-performance gains of ES are only contributed to a sparse subset of larger-magnitude updates. This functional sparsity suggests that large parameter movement need not imply widespread functional change, and held-out evaluations further show that it does not necessarily lead to catastrophic forgetting. Finally, we study how hyperparameter design affects the effectiveness of ES, demonstrating that ES requires a smaller population size in a larger LLM. These findings position ES as a distinct reasoning post-training paradigm rather than a less effective, memory-efficient alternative to GRPO.

As large language models (LLMs) are deployed as autonomous agents, safety failures increasingly involve consequential actions. We study agentic misalignment, where agents take harmful actions under goal conflicts and pressures. Using chain-of-thought (CoT) monitoring, we find that harmful execution is often preceded by intent signals in reasoning. However, post-hoc CoT labels are too coarse to show how intent changes during generation. We introduce INTENT-AS-A-TOOL, an approach that adds intent-targeted tools to give the model a dedicated channel for expressing commitment to a target behavior. The probability of calling an intent tool provides a judge-free, fine-grained signal of the model's tendency to pursue that behavior. Our results show that INTENT-AS-A-TOOL complements CoT monitoring, expands post-hoc CoT labels into dense trajectories, and identifies critical steps for online intervention. These findings suggest that action preferences are useful for tracking agentic misalignment during reasoning. Our code and data are accessible: https://github.com/RebeccaZhang22/intent-as-a-tool.

Recent video generation models are increasingly framed as world models. Many physical processes can unfold in more than one valid way. Therefore, a world model should reproduce not only a plausible trajectory, but also the distribution of possible behaviors under the same initial observation and action. We call this distribution-level requirement probabilistic alignment. However, existing evaluations largely assess individual-video plausibility and do not test whether repeated generations recover the correct distribution. This raises a central question: how far are current video generators from probabilistically aligned world modeling? To answer it, we formalize probabilistic alignment as a distributional criterion for world models and introduce PAWBench, a benchmark for evaluating video generators as stochastic samplers of world dynamics. We further introduce PAWEval, an outcome-level protocol that converts repeated video rollouts into empirical distributions over possible physical behaviors. Across 50 scenarios and eleven current systems, no model consistently matches the reference probabilities while recovering the range of valid behaviors. Having established this gap, we test whether language prompts, initial noise sampling, or model training can reshape the model's predictive distribution. We believe our work can serve as a foundation for future efforts to move towards probabilistically aligned world modeling.

This paper addresses the recovery of essay-scale republication and reuse from fragmented text-reuse evidence, a setting whose central challenge is pair-level evidence consolidation and not fragment retrieval alone. The study focuses on a candidate set centered on essays by eighteenth-century Scottish philosopher David Hume, spanning books from ECCO (Eighteenth Century Collections Online) and historical newspapers. Because the input consists of fragmented reuse hits instead of clean document pairs, and positive coverage is inherently incomplete, we formulate the task as pair-level evidence consolidation into plausible transmission relations and compare three methodological families: a staged rule-based workflow, baselines (a decision tree and two direct LLM settings), and automated rule adaptation. On labeled ECCO--ECCO slices, pair-level feature aggregation alone already reaches 0.948 F1 on the main labeled slice, while the final workflow gives the strongest overall precision-recall trade-off among the tested rule stages. On the full ECCO--ECCO candidate universe, direct LLM baselines flag up to 14,886 pairs as reprints compared to 771 for the final workflow, behaving in this direct-prompt setup as high-recall candidate expanders rather than precision-controlled deployment classifiers. On ECCO--Newspaper, manual audit confirms all 176 predicted positives as genuine cases of republication or reuse, while issue duplication and source-side multiplicity reveal additional provenance structure. Under incomplete ground truth, auditable pair-level evidence consolidation provides a practical way to produce compact candidate spaces for historical inspection.

Steering interventions targeting eval-awareness, a model's recognition that it is being tested, are increasingly used in safety evaluation pipelines, where evaluation-awareness is treated as a single quantity to be suppressed. We show that verbalized eval-awareness in chain-of-thought can be identified as capabilities-flavored ("the user is testing my ability to follow instructions"), safety-flavored ("the user is testing my boundaries"), both, or neither: framings that predict compliance very differently. On Qwen3-32B over the FORTRESS dataset, capabilities-framing predicts compliance with a +24 to +46 percentage-point gap over safety-framing across all tested steering conditions. A CoT-prefill intervention on eval-awareness-negative rollouts suggests the link is causal, with 10 of 11 prefills shifting compliance in the predicted direction. Then, eval-awareness is not behaviorally uniform: aggregate suppression rates can move while the safety-relevant component does not, and the same "X% suppression of eval-awareness" can correspond to qualitatively different behavioral outcomes.

Block drafters propose several tokens in one forward pass, before earlier target tokens are realised. Their rejection mixes two losses: missing within-block path information and imperfect modelling of observable information. Accepted length cannot distinguish them. We separate the two with an information floor, the minimum expected rejection at a specified conditioning order; rejection above this floor is the model gap. Estimating both from target rollouts across four domains, four open-weight targets, and a frontier API target yields three findings. First, the all-parallel floor reaches $0.286$ at the final slot on Qwen3-4B, limiting even the best proposal to $71\%$ per-slot acceptance. Second, one realised token removes $86$--$100\%$ of this floor, a locality also recovered by an independent mutual-information analysis. Third, current drafters remain far above their floors: the final-slot model gap accounts for $43$--$64\%$ of DFlash rejection and $85$--$92\%$ of DSpark's oracle-conditioned rejection. These findings separate the value of short-range conditioning from proposal quality.

Industrial sites contain large volumes of read-only telemetry, but few benchmarks specify how to compile these records into executable multi-turn agent tasks. We present a telemetry-to-episode construction method instantiated as BTS-AgentBench. The pipeline normalizes BTS metadata and raw histories into a read-only tool store, compiles static tasks with tool-derived gold answers and evidence, and lifts retained tasks into typed, bounded operator-facing episodes. The 532-row release adds clarification, goal revision, timestamp policy, quality-gated reporting, and evidence attribution while preserving the source computation and split. Coded contract preflight reports zero findings, and the construction-exclusion controller completes 0/532 rows. Two independent raw-to-episode builds match all 11 logical tool-store exports and reproduce the released 356/87/89 train/dev/test artifact exactly. Applying the shared construction path to XAI4HEAT produces 204 episodes; on its 41-row held-out test split, the controller completes 0 rows and the retained GPT-5.5 execution completes all 41. Code, artifacts, and replay reports are available at https://github.com/kjy7567/BTS-AgentBench.

We establish a global finite-sample guarantee for synchronous quantile temporal-difference learning (QTD) in tabular distributional reinforcement learning. The proof separates two stability mechanisms. A global comparison argument, based on the order monotonicity of reward cumulative distribution functions and the $W_\infty$ contraction of the distributional Bellman operator, brings an arbitrarily initialized iterate into a local neighborhood. Inside that neighborhood, we linearize the QTD mean field. Its Jacobian is a nonsingular $M$-matrix, and the associated positive semigroup permits a variance-sensitive martingale analysis. For stepsizes $α_t=c(t+1)^{-a}$ with $a\in(1/2,1)$, the leading last-iterate fluctuation is of order $\widetilde O\bigl(T^{-a/2}/\sqrt{1-γ}\bigr)$ and has no polynomial dependence on the number of quantiles. The deterministic transient and the required burn-in can still depend on the smallest Bellman-target density, which is of order $m^{-1}$ in the worst case. The result therefore distinguishes sharply between the local stochastic fluctuation and the global sample complexity.

Agent harnesses shape how language-model agents use instructions, tools, and runtime components, but adapting these harnesses requires costly verification. Existing propose-and-verify methods typically score every candidate on a fixed task set, wasting rollouts on unrelated behaviors and allowing aggregate scores to obscure specific regressions. We introduce HarnessLens, a budget-aware framework for automated harness evolution. HarnessLens jointly explores the task space and user-configurable components, derives candidate modifications from execution trajectories, and selectively verifies each candidate on behavior-relevant tasks using an attributable-evidence gate. Across three agent harnesses and four benchmarks, HarnessLens improves average held-out performance by 7.6-13.6% while consuming substantially less evaluation budget than competing baselines. These results demonstrate that behavior-aware verification with explicit attribution enables more reliable and sample-efficient harness evolution under constrained interaction budgets. Our code is available at https://github.com/jhxu5214/HarnessLens.

Audits of LLM judges certify a bias by contrasting matched conditions, and the strongest designs difference twice: a within-item contrast between two candidate responses, differenced again across a manipulated attribute, read off a bounded rating scale. We show that this endpoint is not identified on the scale that reports it. Each term of the double difference is censored by its own share, so the observed statistic confounds differential preference with differential attenuation: a severity shift common to both responses manufactures an interaction whenever the two censor it unequally, as unequal distances from the bounds make them, exactly where good stimuli place them. We exhibit the failure inside a pre-registered audit of a frozen pedagogy judge, sealed before the first of its 990 calls. The registered primary endpoint, the effect of a stated learner profile on the judge's scaffolding preference, is null: $+0.085$ points (95\% BCa $[-0.167, +0.353]$, $p = 0.684$). The audit's one nominally significant interaction, $+0.378$ ($p = 0.002$), is not identified as preference: a construction containing zero differential preference reproduces 79 to 85\% of it from the observed severity shift and the scale floor alone. We derive the mechanism in closed form and show that its contribution is measurable from an audit's own ratings.

Accurate identification of the correct view or angle in cardiac ultrasound (echocardiogram) is a critical component of cardiologic imaging. This step is essential for precise anatomical interpretation, reliable measurement, and the reduction of clinical errors. Although computer vision has advanced significantly, most state-of-the-art models perform well on standard benchmarks but often yield suboptimal results in specialized medical imaging tasks due to the high level of noise present in the data. QuantumBoostNet, a hybrid classical-quantum architecture, is introduced to address these challenges. This model integrates a classical backbone with two heads: one classical and one quantum, with the quantum head implemented as a parametrized 10-qubit quantum circuit. Training occurs in two stages, with an adaptive transition between heads governed by a mixing parameter that monitors loss dynamics. Extensive experiments indicate that, despite the limited number of qubits that can be simulated, QuantumBoostNet consistently outperforms state-of-the-art classical and hybrid classical-quantum models in cardiac ultrasound view identification, achieving a relative improvement over the best competitor. QuantumBoostNet also demonstrates superior performance on established image classification benchmarks and exhibits robustness to noise. These findings support the continued development of hybrid classical-quantum models for specialized medical imaging applications.

Instruction hierarchy is a model-side defense that assigns instructions different levels of privilege according to their sources. These levels constrain which content may direct model behavior. During agent execution, however, agent harnesses construct context for each model invocation. This construction can elevate low-level content to a higher instruction level and grant it greater model-facing privilege. We introduce instruction privilege escalation. In this attack, an attacker induces an agent to elevate low-level malicious content to a higher instruction level. The elevated content then causes the agent to execute instructions it would not follow at their original level. We evaluate this threat by using multi-agent mechanisms to achieve 13 attack objectives across six coding-agent harnesses. These objectives span confidentiality, integrity, availability, and remote code execution. With unrestricted action execution, the attacks achieve all 13 objectives on all six harnesses. Under automatic permission review, the attacks achieve all 13 objectives on all three harnesses that provide this mode. We further reproduce the vulnerability using harness-provided persistent goals and scheduled tasks. These results demonstrate the generality of instruction privilege escalation.

We ask whether large language models (LLMs) can design effective algorithms for well-specified operations research (OR) problems. We study inventory control, queueing network control, and assortment optimization. We evaluate two levels of LLM use: at level 1, the model receives one problem instance and returns a solution for that instance; at level 2, it receives only the problem class description and broad parameter ranges, and returns an algorithm that maps instance parameters to solutions. Human input is minimal: we give one untuned prompt that describes the problem, and the model has access to a Python sandbox tool with a fixed compute budget. The strongest model we test, gpt-5.6-sol, matches or outperforms the best existing method on almost all evaluated instances. This holds even at level 2, where the returned algorithm is fixed before seeing the evaluation instances. Performance also improves sharply across models released less than eight months apart, suggesting that this capability is moving quickly. Thus, for the well-specified operations problems we study, a single untuned LLM query can already produce algorithms competitive with specialized methods. These results suggest that frontier LLMs can be a serious empirical baseline for algorithm design in well-specified OR problems.

Music recommendation relies primarily on two signals: user-item interactions, which fail in the cold-start regime, and intrinsic musical content, available for any recording. We argue that a third, largely untapped signal is both richer and more principled: critical adjacency, the pairwise relation established when an expert critic explicitly links two artists in long-form prose. It encodes deliberate judgments about which artists belong together. Prior work established its internal validity, showing it recovers coherent, interpretable communities and can match collaborative filtering in user-satisfaction simulations, with no user data. What has been missing is external validation: whether this critic-sourced relation is grounded in the music itself versus sociological context. We test it against acoustic content, reframing the question as one of construct validity. Representing artists as empirical distributions over 80 low-level Essentia acoustic descriptors and modeling pairwise proximity via marginal optimal-transport (Wasserstein) distances, we evaluate how far critical adjacency is sonically recoverable under a cold-start, artist-disjoint split. Our ensemble recovers these edges at out-of-sample AUC of 0.767 (95% CI 0.761-0.775). Recoverability rises monotonically with critical consensus, reaching 0.865 on multi-source attested edges. Stratified evaluations align with sociological models of genre: tightly bounded, scene-based genres show higher recoverability than broad industry umbrella terms. Critical discourse is thus a rich source of information for recommendation, decomposing into a reproducible "sonic core" and a "sociological remainder" driven by narrative positioning, subcultural context, and canonical placement. The work offers both a scalable cold-start discovery mechanism and a sociologically grounded approach to MIR and MRS research.

Inferring molecular structures from multimodal spectroscopic measurements requires integrating complementary yet highly heterogeneous signals. However, the common paradigm of directly concatenating multispectral sequences can exhibit anomalous performance degradation, primarily due to pronounced heterogeneity and the resulting multimodal imbalance across modalities. As a remedy, we propose MM-Spectrum, a sparse Mixture-of-Experts framework tailored for multimodal multispectral spectra-to-structure elucidation. To better match the information characteristics under multispectral imbalance, MM-Spectrum introduces an explicit modality-aware routing mechanism that exposes spectral identity to the router in addition to token content representations. Moreover, it incorporates shared and interaction experts, together with heterogeneous expert capacities, to extract multispectral modality-unique and cross-modal synergistic information while suppressing noise-induced interference. Across full-modality, bimodal, and missing-modality settings on molecular structural elucidation, MM-Spectrum achieves consistent and substantial improvements, supported by ablation studies and interpretability analyses.

Most single-stage 3D object detectors complete different tasks with the same extracted features. Nevertheless, it is impossible to project features into a common space that is adaptive for all the tasks. We present a novel task-aware deformable prediction (TADP) method for single-stage 3D object detection to solve this problem. Firstly, a triple feature refinement aggregation module is designed to extract three-level features adaptively. Additionally, we design the multi-scale feature aggregation block to fuse multi-scale features in a scale-aware manner. Finally, the prediction of each task is deformed with the designed plug-and-play task-aware deformation head. It can percept the emphasis and interaction of each task. We also designed three different deformation modules. The experimental results demonstrate that the proposed deformation head shows good results on other detection methods. The experimental results on the KITTI dataset demonstrate that the car mAP is 80.91%, surpassing many state-of-the-art methods on the KITTI benchmark.

Although Large language models (LLMs) mediate access to knowledge and computational assistance, their capabilities should benefit vulnerable groups in the same way. However, it is unclear whether existing AI systems are inclusive enough for blind and deafblind users to access the same functionality through Braille, whose indicators, contractions, and digital representations introduce distinct requirements for model comprehension. To this end, we introduce BrailleBench, a benchmark for evaluating LLMs in Braille comprehension from different Criteria. BrailleBench aligns 5,570 instances from five datasets, including mathematics, commonsense, and multi-hop question answering across English and Braille Grades 1 and 2. Different configurations are designed to understand whether the systems can comprehend Braille-authored content, express answers in Braille, and complete end-to-end Braille interaction. To ensure the quality and prevent evaluation bias, the benchmark is built through a deterministic, expert-reviewed pipeline via a self-created Braille Toolkit without using any data instances generated by LLMs. We evaluate six representative LLMs from various aspects. The results reveal a persistent gap between print-English capability and Braille accessibility. Braille understanding and expression are asymmetric, where Grade 2 is especially fragile on the input side compared to Grade 1, and fully Braille requests further reduce performance. The experimental observations provide valuable guidance for the development of future Braille AI systems. All related resources in BrailleBench are publicly available for future research.

Efficiently improving autonomous agents across diverse tasks is central to accelerating recursive self-improvement (RSI) in agentic AI, with prompt optimization emerging as a promising approach capable of delivering performance gains comparable to those achieved by fine-tuning model weights, while reducing computational costs in both optimization and serving. However, recent developments increasingly favor unnecessarily complex prompt optimizers. We introduce Naive Prompt Optimization (NPO), a lightweight single-lineage method that iteratively revises prompts using a teacher model with rollout feedback. NPO achieves comparable or better performance than GEPA with fewer rollouts, and its advantage increases with stronger teacher models, suggesting that stronger teacher reasoning can partially substitute for optimizer-side search complexity. In interactive games, NPO remains broadly competitive with GEPA, while GRPO performs better on some tasks less amenable to prompt optimization. We also show that NPO-optimized prompts elicit similar performance improvements when applied verbatim to other student models, especially across models within the same family. Overall, our preliminary results show that simple, linear prompt optimization can rival substantially more sophisticated and complex search procedures.

Latent chain-of-thought models move intermediate reasoning from emitted text into continuous states, improving compactness but hiding the causal object. We introduce SCIT, the Suffix Cache Interchange Test, a causal protocol that constructs exact source-recipient counterfactuals, patches declared cache segments, and identifies which transformer object carries the counterfactual computation. SCIT combines sufficiency tests with K/V component splits, hidden-state controls, semantic source controls, decoded validation, and matched corruption. On CODI-GPT2 and a Sim-CoT-style GPT-2 reproduction, counterfactual arithmetic transfers primarily through value-cache suffix trajectories rather than hidden states, keys, reusable answer slots, or single-token triggers. Complete sufficiency-and-necessity evidence for the late-value-suffix mechanism holds for the main CODI-GPT2 checkpoint; the Sim-CoT-style checkpoint shows the same sufficiency and decoded-control pattern but insufficient matched-corruption evidence for a necessity call. Beyond these local arithmetic cells, SCIT reveals carrier-regime shifts: arithmetic-like GPT-2/1B cells preserve latent-tail value/KV transfer, whereas competent 8B and repaired non-arithmetic cells route through prompt-prefix or full-cache K/V; boundary cells receive no mechanism call. SCIT therefore contributes a cache-level diagnostic, a checkpoint-specific GPT-2 arithmetic mechanism, and a competence-gated carrier map rather than a universal latent-tail claim.

LLM agents increasingly rely on generated interaction data to learn how to interact with external environments. Agentic data generation must maintain consistency among environments, tasks, interactions, and success signals while producing experience that is useful rather than merely abundant. Existing work spans many agent domains, but domain-centered organization and heterogeneous evaluation often obscure common generation mechanisms and conflate candidate construction with verification and selection. This work develops a two-level framework for the field. First, we represent agentic data as a common factorized object $(E,q,τ,v)$, comprising an environment specification, task signal, interaction realization, and optional verifier. We organize generation paradigms by their primary anchor and dependency structure. Second, we formulate generation as constrained distribution design through the Accuracy-Complexity-divErsity (ACE) lens. Accuracy establishes the feasible support of grounded and internally consistent data. Within this support, Complexity places learning mass relative to the capability of a declared learner and execution configuration, while divErsity controls coverage and redundancy of data. Using this framework, we explore how prior work verifies generated experience, constructs and calibrates difficulty, and expands behavioral coverage. The literature reveals a shift toward execution-grounded accuracy, learner-relative complexity, and diversity beyond surface variation or dataset size. We further discuss broader directions and emerging trends in agentic data generation through the ACE lens, including their implications for scaling, data sources, training regimes and adaptive learning. Overall, the central challenge is not simply to generate more data, but to continually allocate valid, informative, and non-redundant experience as agents and environments evolve.

World Action Models (WAMs) augment robot policies by predicting how task-relevant scene states may evolve under interaction. Recent WAMs increasingly perform such prediction in latent representation spaces, avoiding full appearance-level generation while preserving control-relevant information. Yet latent transitions are commonly realized with Transformer-based predictors whose inductive structure is centered on token interaction rather than temporal evolution. We study transition realization as an architectural choice distinct from predictive representation and prediction-policy coupling. We introduce the Latent Evolution Operator Network (LEON), which models latent evolution in a learned observable space through context-modulated operator-based propagation and additive forcing. Grounded in the controlled Koopman generator view of evolution, LEON organizes context-dependent transition variation around a shared evolution-operator structure while retaining a complementary path for additive change. Controlled dynamical systems verify the resulting evolution-specific inductive bias and the complementary roles of operator propagation and forcing. Across two WAM formulations that integrate latent prediction into the policy differently, LEON improves closed-loop performance and robustness while remaining effective under full transition replacement. These results establish transition realization as a consequential architectural choice in latent WAMs.

Operator learning in scientific machine learning is concerned with approximation of maps between infinite-dimensional function spaces; such maps frequently arise as the solution operators of partial differential equations (PDEs). Neural operators have demonstrated broad empirical success at approximating such maps from data. However, most existing neural operator architectures enforce boundary conditions indirectly through training from data even though the boundary condition is often known exactly. Furthermore, existing modifications and approaches that do enforce boundary conditions explicitly suffer from impractical restrictions, including boundary smoothness, uniform grids, and separable, box-like domains. In this work, we propose an architecture which, independently of training, satisfies homogeneous Dirichlet boundary conditions, whilst simultaneously retaining the expressivity of existing kernel-integral neural operator architectures. This is achieved by enforcing the property that the output of each layer is contained in the span of a subset of the homogeneous Dirichlet eigenfunctions of the Laplacian on the output domain. The method requires only that the output domain be bounded with Lipschitz boundary and places no restriction on the choice of discretization, making it applicable to arbitrary mesh data and general geometries. We prove universal approximation for the resulting architecture; furthermore the approach we adopt in the analysis proves universality for a broad class of kernel-integral neural operators thereby uniting existing theory for a variety of operator learning methods. We validate the proposed method on maps defined by the coefficient to solution map in 2D PDEs: Darcy flow on a square domain and the Helmholtz equation on a circular domain. Comparisons are made with alternative methods.

One approach to mechanistic interpretability explains behavior through circuits: the components and connections that carry it. Frozen discovery often returns hundreds of edges, making them hard to inspect, compare, or verify exhaustively. We introduce Circuit Condensation, which post-trains models to concentrate behaviors into smaller causal graphs. Each round prunes low-attribution edges and trains a low-rank adapter to match the original through what remains, retaining the cut only if task performance and general capability survive. Across four behaviors and eight models, condensed circuits are smaller than the strongest frozen baseline in 30 of 32 settings, by $8.1\times$ on average and up to $316\times$. Repeating the search without weight updates produces larger circuits in 29 of 32 settings, showing that weight updates, rather than search alone, drive the reduction. Testing every subset of 19 circuits finds 11 that cannot be reduced and reveals removable edges in the rest. Pair ablations expose dependencies between edges, showing that their effects cannot be understood independently. On indirect object identification, condensation isolates 24 heads, 17 of them with documented roles, against 61 heads and 36 undocumented ones for the matched frozen circuit: a sufficient sub-circuit of the published mechanism rather than a reconstruction of it. The resulting circuit tracks the original model's next-token distribution and predicts its errors.

An LLM call in a semantic data processing system is expensive enough to dominate query cost, yet slow enough to hide a CPU-side learner's update behind its round-trip. In production, LLM compute accounts for $80-90\%$ of query cost, and each call costs $10^5-10^7\times$ a relational predicate. The latency window inverts a design constraint of classical adaptive query processing, where online learners had to stay lightweight to avoid dominating the predicates they optimize. At LLM latency, per-call gradient steps and per-batch threshold solves fit inside the round-trip. We develop compositional online learning at the LLM call boundary: a framework for combining online-learning components in semantic data processing systems. Each component makes execution-time decisions and refines its learned artifacts online. The design space spans two axes, decision granularity and learner update cadence, and the components share a single learning pattern that hides each trainer step inside the next LLM round-trip. A production case study in Cortex AISQL composes three components: a memoization layer, an online per-call filter-ordering learner, and an online per-batch cascade-routing learner. A conditional cost decomposition assigns each learning component to a distinct factor of per-row LLM cost. Under independence, the two learning components compose multiplicatively to an $11.4\times$ upper bound on a representative conjunction-filter workload. Self-selection at the cascade boundary, sample-budget shrinkage, and selectivity-estimation drift reduce it to a realistic figure near $8\times$.

Computationally demanding and opaque deep learning models can be better understood and optimized by analyzing how they transform data. While deep transformers have been widely studied in computer vision and natural language processing, their application in tabular data remains relatively underexplored. This paper presents one of the first applications of an importance-scoring metric to interpret multi-head transformer models in learning from tabular data. Experiments conducted on 40 diverse tabular datasets demonstrate robustness to head drops based on the proposed head importance score. In 72.5\% of experimental examples, the model remains most resilient to performance drops when heads with the lowest importance scores are gradually removed. In contrast, removing the most important attention head first results in the greatest reduction in classification performance. A closer look at individual head importance scores across six attention layers reveals that important heads are scattered across layers, with no consistent layer-specific trends. In contrast to the image and language domains, the importance of individual attention heads varies considerably across tabular datasets with different schemas and feature spaces. The proposed importance score can improve efficiency and redundancy within transformer architectures. We make the source code for measuring the importance of individual attention heads publicly available.

Adverse drug reactions (ADRs) are a major, largely preventable source of patient harm. In high-income settings, electronic health records store a patient's allergy history and warn prescribers when a contraindicated drug is ordered; in rural Bangladeshi public hospitals no such record exists for outgoing patients, a single physician may see on the order of one patient per minute, and a patient's history of severe reactions does not survive between visits. This paper proposes and outlines the evaluation of a lightweight, smartphone-based safety-check system for this setting. At registration a soft identifier (a phone number) is recorded; after the physician writes a prescription, its image is captured, the brand names are resolved to active ingredients using national drug references, and the ingredients are matched against the patient's recorded severe reaction history. The system is retrieval-based rather than predictive, and is silent by default, raising a flag only for high-risk matches a design grounded in the alert-fatigue literature. We frame the work as a feasibility study: we describe the proposed framework and an evaluation plan measuring workflow fit under high volume, usability, identity-resolution reliability, and retrospective detection of known reaction cases. We explicitly do not claim a clinical-outcome effect, which the low base rate of severe events places beyond a single-site feasibility study.

Large language model (LLM) agents increasingly operate over untrusted webpages, documents, tools, and persistent states while exercising authority over security-sensitive resources. Existing defenses typically protect either planning or individual tool interactions, but persistent agents face a broader threat: attacker-controlled data can alter control flow, enter security-sensitive tool arguments, or compromise later queries. We present SPA, a plan-first architecture that secures planning, execution, and cross-query state reuse. SPA invokes the planner once per query to generate a complete executable plan in a declarative domain-specific language, then applies dual-lattice information-flow control to track confidentiality and integrity across explicit data flows and control dependencies. To support persistence without re-exposing untrusted payloads to the planner, SPA stores execution results as labeled artifacts and reveals only semantic metadata during later planning. We evaluate SPA on AgentDojo and AgentDojo-MQ, which is our multi-query extension for measuring secure state reuse and delayed attacks. Under the 'tool_knowledge' attack, SPA with information-flow control reduces attack success to zero on AgentDojo and 0.2% on AgentDojo-MQ. Our results show that plan-first execution combined with label-preserving persistence can substantially strengthen persistent LLM agents, while revealing an important security-utility tradeoff introduced by strict integrity enforcement.

Human Activity Recognition (HAR) using inertial measurement units (IMUs) enables a wide range of applications, yet the field still lacks a unified model that can generalize across diverse subjects, devices, and activities. Training such a model is difficult due to two key challenges: sensing heterogeneity -- differences in sampling rates, channel configurations, and sensor placements -- and poor generalization to unseen activities and label vocabularies. We introduce HALO (Heterogeneity-Aware Language-aligned Open-set model), a domain-specific IMU foundation model that addresses both challenges through a two-stage training framework. Stage 1 pretrains the IMU encoder with heterogeneity-aware self-supervised learning, including adaptive-pooling tokenization, channel-independent feature extraction, and contextualized sensor conditioning that injects natural-language sensor descriptions into each channel embedding. Stage 2 aligns this IMU encoder with text embeddings via synonym-aware soft contrastive learning, enabling open-set recognition via cosine-similarity retrieval without per-dataset classifiers. Trained on 10 public HAR datasets and evaluated on 7 held-out datasets, HALO outperforms five state-of-the-art baselines on all 8 aggregate metrics, and still leads on 3 of 4 settings under baseline-matched inputs. Despite using only ~35M trainable parameters -- 10x fewer than the latest foundation model MOMENT (341.2M) -- HALO improves zero-shot open-set accuracy, measured over all 87 training labels, by 13.7 percentage points. On two further datasets with severe distribution shift, every model including HALO collapses zero-shot. A video demonstration of HALO's performance in real world is available at https://youtu.be/rooVKragtFU

Effective human-robot collaboration in industrial settings requires robots to understand human intentions and assist with task planning, reducing workload. Recent works have explored the use of Multi-modal Large Language Models (MM-LLMs) for task planning in such data-scarce scenarios, leveraging in-context learning to interpret user actions and generate long-horizon action plans in natural language. However, MM-LLMs inherently lack an understanding of system states and do not track state transitions, often leading to hallucinated actions that deviate from the intended goal. Additionally, generating action plans in natural language tends to limit the generated plans to a high level, introducing ambiguity in action execution. To address these limitations, we propose the State-aware Task Estimator and Planner (STEP), which prompts a MM-LLM to explicitly estimate the state of the system and predict the state transitions resulting from executed actions. By forecasting future states alongside actions, STEP ensures task-convergent planning while also providing additional assistance parameters necessary for executing the predicted actions. We evaluate STEP in a simulated environment using a robot assembly task. Our approach outperforms the state-of-the-art by 32.8% in action executability and 14.8% in final-state error.

Mental health assessment relies on episodic self-report scales, which convert subjective states such as stress into numerical scores but provide only sparse snapshots of wellbeing. Wearable devices offer longitudinal behavioral and physiological signals for continuous, low-burden monitoring. Recent LLM-driven personal-health agents enable natural language queries over wearable signals, but mainly handle short-term, retrieval-based lookups (e.g., highest step count over a week). They do not evaluate whether agents can reason over long-term signals to predict wellbeing scores paired with evidence-grounded rationales. To address this gap, we introduce BALMS, the first systematic benchmark of LLM-based agentic systems for longitudinal mental health sensing. BALMS spans 3 real-world longitudinal datasets, 2 task families (closed-form wellbeing-score prediction and rationale generation auto-graded by an LLM-as-Judge), 3 agentic paradigms evaluated across 5 open- and closed-source LLM backbones. We find that zero-shot agents rarely outperform a simple mean baseline, except with stronger backbones or compact, semantically meaningful features. Chain-of-thought prompting improves reasoning-oriented backbones, but does not guarantee temporal grounding or numerical correctness. Together with more analysis on efficiency and temporal scaling, BALMS highlights the need for longitudinal mental health agents that selectively retrieve history, ground temporal evidence, and reason over interpretable behavioral features.

Vision-Language Models (VLMs) demonstrate exceptional visual reasoning capabilities, yet their inference costs escalate rapidly with the proliferation of visual tokens. Existing visual token pruning methods exhibit two fundamental limitations. First, most approaches operate exclusively post-vision encoder, leaving the substantial latency of the visual encoding phase unoptimized. Second, under strict token budgets, these methods often fail to jointly preserve holistic visual contexts and fine-grained details, leading to performance degradation. To address these bottlenecks, we propose PACE (Pixel-Adaptive Condense and Extract), a training-free inference framework that accelerates both the vision encoder and the Large Language Model (LLM) via a unified Condense-and-Extract paradigm. During the Condense stage, an Adaptive Pixel Compressor (APC) evaluates visual information density prior to encoding, adaptively downsampling redundant inputs, curtailing encoder computation while preserving global context and essential visual cues. In the Extract stage, a Dynamic Dual-Attention Extractor (DDAE) selectively retains visual tokens via a fusion of internal visual signals from the encoder and semantic signals from the LLM, safeguarding task-critical details. By integrating PACE into Qwen2.5-VL-7B, the model retains 93.8% of its original performance while utilizing only 10% of the visual tokens, yielding a 3.1x speedup in time to first token (TTFT). Our code is available at https://github.com/jjL357/PACE.

Nitrogen rates for winter wheat are set before the season, under unknown prices and weather. The standard UK advice does not respond to prices, yet recent price swings moved the most profitable rate by tens of kilograms per hectare. Machine learning is often proposed as the fix. However, it is usually judged on prediction accuracy, and accurate prediction does not by itself make the recommended rate more profitable. Our insight is to score nitrogen advice directly by the profit it forgoes on measured yield response curves. We build a test bench on 892 such curves from two long running UK experiments, and sweep the nitrogen to grain price ratio to cover all price scenarios. On this bench, machine learning fails as a predictor. No model recovers the best rate within farm tolerance, and the benchmark noise shows none can. At normal prices, every model also loses to the standard advice on profit. The gain sits elsewhere. A simple correction step applied after the model cuts profit losses by a quarter, while better models and extra features give no gain. The same frozen correction cuts losses by 43% at the second site without any retraining. A hybrid of standard advice plus a damped correction removes bias and trims rare large losses. The same price sweep also prices emission cuts, at a cost comparable to current carbon prices. Machine learning therefore pays as a profit scored correction to standard advice, not as its replacement.

A known necessary condition for Fisher consistency of the structured support vector machine requires the task loss to be a metric for which every output triple has a common geodesic point. We show that this condition is not sufficient for the canonical coordinate-wise argmax decoder. A four-output unit star admits an exactly optimal score vector whose maximizers are all strictly non-Bayes, and four outputs are minimal among metrics satisfying the condition. We then completely classify positively weighted tree metrics whose vertex set is the output space: argmax consistency holds if and only if the tree is a path. The failure on branching trees is confined to boundary distributions; every tree retains the argmax property at every full-support distribution. Among metrics satisfying the common-geodesic condition, five outputs are necessary and sufficient for a full-support counterexample; $K_{2,3}$ is the smallest member of an infinite $K_{m,n}$ family. We additionally give a full-support counterexample for the three-dimensional Hamming cube. All optimality claims have exact primal-dual certificates. The counterexamples expose a concrete decoder gap: in this polyhedral setting, an embedding can guarantee the existence of a calibrated link without validating a prescribed argmax link on every surrogate-risk minimizer.

In 2025, the US government launched an unprecedented series of attacks on its own scientific research groups. A year later GitHub dropped below 90% availability for the first time, while wildfires in Canada, France, Spain, and elsewhere forced researchers from the homes and labs. These events and others have reminded us just how fragile research computing systems can be, and that planning for disasters is one of the most effective ways to prevent them. This paper is a short guide to disaster planning and recovery for a small research software team. The tips assume you are doing everything yourself on top of your regular job, and that you aren't an experienced system administrator. Some of the tips do require that kind of expertise, but most research institutions have research computing groups, data librarians, and environmental health-and-safety offices whose entire job is to help with exactly these problems. This paper tells you what "done" looks like; they can often provide it.

Peer effects are difficult to estimate when interaction graphs evolve because pre-assignment network history, dynamic peer exposure, and post-assignment network change have distinct causal roles. We introduce a controlled contrast framework that indexes potential outcomes by own treatment, temporally aggregated peer exposure, and a post-assignment evolution summary. Differences between the resulting means define own-treatment, peer-exposure, controlled network-evolution, and joint controlled contrasts rather than a mediation decomposition. We develop the Dynamic Network Doubly Robust estimator, DynaNet-DR, which combines a temporally factorized propensity with normalized augmentation. Under consistency, summary sufficiency, sequential exchangeability, positivity, nuisance convergence, and weak dependence, its canonical estimator is consistent when either the outcome regression or the propensity estimator is consistent. The reported implementation adds representative-score prediction, fixed clipping, and finite-sample stabilization. Semi-synthetic benchmarks on fixed real temporal graph sequences show favorable estimation accuracy among methods targeting the full profile. These benchmarks assess summary-indexed contrasts rather than counterfactual edge generation, and the MathOverflow study is an observational illustration under the stated assumptions.

In closed cattle-fattening barns, the indoor climate and herd growth are mutually interdependent. Temperature, relative humidity, airflow, and ventilation affect thermal comfort, feed intake, metabolic heat production, daily growth, feed efficiency, and energy consumption, while body-weight gain alters the future heat and moisture loads of the barn and, consequently, its ventilation, heating, and energy requirements. This article proposes a trans-domain digital twin framework with single-episode learning capability, customized for bio-aware climate and energy control in a closed cattle-fattening barn. The framework integrates a mechanistic climate simulator, a livestock growth simulator, model predictive control, lightweight reinforcement learning, and structured knowledge memory within a multi-rate temporal-loop architecture. The fast temporal loop operates every five minutes to evaluate actuator decisions and maintain short-term thermal comfort, safety, and energy efficiency, whereas the slow temporal loop provides biological guidance based on daily climatic conditions, feed efficiency, heat production, and growth-limiting factors. The results show that climate, growth, energy, feed, biological guidance, and memory can be linked within a single executable control cycle. Remaining limitations include the need for field validation, improved management of feed pressure, and reduction of abrupt actuator-command variations.

Electronic design automation (EDA) has advanced engineering productivity through successive generations of tooling that progressively automate synthesis, optimisation, and verification. Large language models (LLMs) extend this trajectory by enabling direct translation from design intent to hardware implementations. In most of the EDA literature, LLM-based solutions are typically assisting siloed design stages or tasks, however this obscured the drivers by which capability emerges and systems scale. In this Perspective, we instead define three hierarchical roles that reveal how capability accumulates: a Generator that produces design artifacts in a single pass, an Agent that refines outputs through iterative tool feedback, and an Orchestrator that coordinates decisions across EDA-stages. Across published systems, this reveals a syntax trap in which models are trained to produce plausible code rather than physically correct hardware, compounded by fragmented tools and loss of design context that obscure how decisions affect later stages. Comparisons across the three roles show that current approaches struggle to scale to industrial designs, motivating a shift towards a standardised, physics-aware orchestrator that connects tools and agents across the EDA flow for more reliable and accessible hardware design.

LLM agents are increasingly applied to anomaly detection and root-cause analysis in time-series observations collected from real-world systems; however, their performance on these tasks has not been systematically evaluated under controlled conditions. We introduce TraceBench, a simulation-based framework for generating controlled root-cause attribution tasks. In each generated task, an agent receives time-series observations produced by simulating a physical dynamical system and must determine whether a system parameter was altered during the simulation and, if so, which one. Using TraceBench, we generate tasks from three interpretable mechanical systems and systematically evaluate four LLM agents across controlled experimental conditions, yielding new insights into how these agents analyze time-series observations from dynamical systems. Our results show that agents benefit substantially from domain context and explore data primarily through numerical console output rather than visualizations. We also find that agents generally perform worse when required to produce a Python script that maps each time-series sample to a predicted root-cause label than when they submit predictions directly. We release our datasets, agent trajectories, experimental results, and a leaderboard on our website, tracebench.github.io.

Understanding spoken dialogue requires joint reasoning over lexical content and paralinguistic acoustic signals such as emotion and conversational intent. However, existing evaluations often allow shortcuts based on transcripts or single-modality solutions, obscuring whether models genuinely ground predictions in speech. We formalize this failure mode as cross-modal disagreement, where transcripts suggest plausible but incorrect surface interpretations while acoustic cues such as prosody or speaking style support different answers. We develop a scalable framework that identifies text-biased surface interpretations and converts disagreement regions into conflict QA examples. We also include consistent cases where transcript-based and speech-grounded interpretations agree, enabling evaluation beyond adversarial audio dependence. This results in ContraTalk, a controlled benchmark containing 501 questions across five discourse dimensions: interaction behavior, emotion state, dialogue act, social stance, and conversational intent. We further develop an agentic-style reasoning framework that converts speech into an Audio Twin, a text-readable representation of localized acoustic cues that exposes acoustic evidence to the reasoning model. Experiments show that strong text-only LLMs exceed 90% accuracy in consistent cases but drop to 33-48% in conflict cases. Direct AudioLLMs provide only partial grounding, still selecting the transcript-biased trap in roughly 30-40% of conflict cases. Our Audio Twin framework improves conflict-case accuracy while reducing trap selection, but its consistent-case behavior remains backbone-dependent. These results identify transcript-based shortcuts as an important failure mode in spoken dialogue understanding and show that explicit acoustic evidence aggregation provides a more controllable interface for diagnosing and improving speech-grounded reasoning.

An LLM agent shown a professional-looking market panel commits to a directional call on a provably unpredictable question far more often than one asked the bare question: across 12 frontier models, commitment rises from 6.5% to 54.0% as evidence is escalated. It commits just as readily when every number on the panel is invented: fabricating the entire display, so nothing the model can see is true except the question itself, still lifts commitment from 24.5% to 36.8%, statistically indistinguishable from the 37.6% produced by genuine market data. What unlocks confident action is not information but the authority of its packaging. The failure is narrow and locatable. Incapacity is not the answer: on matched answerable questions attached to the same panels, the same models answer essentially always, at near-perfect accuracy. Nor is it belief - stated probabilities barely move across the gradient that swings action by 48 points, and score worse than a climatological baseline. Missing judgment isn't it either: asked to classify a question's knowability before acting, models call it irreducible 90% of the time and then commit on just 0.4% of those. The act/don't-act gate is what fails, and the effect is concentrated in a few models rather than universal. Because the gate is separable, it can be trained. Supervised fine-tuning of a 3B model on 540 synthetic cases, predominantly dice, coins, jars and timers, drives commitment to 0.0% on the original cases and transfers to three unseen domains. It does not survive everything: the gate holds exactly when the response format leaves room to reason, and rigid formats that remove that room leave the model confident and wrong on questions it otherwise answers correctly. The gate is trainable and context-fragile, and deployment needs both halves of that sentence.

Autoregressive large language models (LLMs) routinely generate factually incorrect outputs with high decoding confidence, limiting their deployment in high-stakes workflows. Existing output-stage uncertainty metrics can fail when models are overconfident on false assertions, while multi-sample verification pipelines introduce substantial memory and latency overhead. This work evaluates whether internal hidden-state transition dynamics during generation can signal factual errors without auxiliary decoding calls. We introduce Prediction of Prediction (PoP), a mechanism that captures layer-transition uncertainty by fusing intermediate hidden representations across depth during a single forward pass. Evaluated on the TruthfulQA benchmark using autoregressive transformer backbones, PoP achieves an area under the receiver operating characteristic curve (AUROC) of 75.5% for factual-correctness classification. The mechanism operates within the base forward pass, adding less than 1.2% runtime latency and requiring zero additional generation passes. The numerical results are reported from the author-verified experimental implementation and are bounded by the evaluation scope described below.

Battery lifetime is central to sustainable electrification, yet the particle cracking that drives lithium-ion cathode aging is hard to measure: quantitative microscopy of this degradation is bottlenecked by annotation, because each destructive electron-microscopy cross-section spans hundreds of megapixels and pixel-level expert labelling requires hours per image. We show that a frozen self-supervised vision-transformer encoder, combined with a lightweight trainable decoder and iterative model-assisted annotation, turns this sparse labelling budget into population-scale degradation measurements. Applied to three 120-megapixel NMC cathode cross-sections representing initial, cycled-aged and calendar-aged states, the framework distinguishes intragranular cracks from early- and late-stage intergranular cracks and yields per-particle distributions of crack width, tortuosity and area fraction. Late intergranular crack coverage reaches 4.6% in the cycled sample versus 0.5% in the initial and calendar-aged samples, forming more tortuous, higher-coverage networks, consistent with degradation from repeated electrochemical cycling rather than elevated-temperature storage alone. A single destructive image yields the population-level statistics needed for lifetime-extending design, aging assessment and second-life decisions.

Large Language Models (LLMs) have enabled a shift from sentence-level to document-to-document (Doc2Doc) machine translation, promising improved global coherence. However, document-to-document generation in a single pass frequently suffers from structural misalignment, manifesting as sentence omissions or hallucinations that violate the core requirement of source-target correspondence. To address this, we introduce Sentence Translation Alignment Rate (STAR), an auxiliary metric that explicitly quantifies sentence-level structural fidelity. Building on this, we propose STAR-masked Preference Optimization (StarPO), a framework that ranks document-level hypotheses by structural quality and utilizes a dynamic alignment mask to focus optimization on misaligned segments. Experimental results across news and literary domains demonstrate that StarPO significantly enhances translation quality and structural integrity. Notably, StarPO allows compact models to surpass the performance of massive proprietary systems like GPT-4o while maintaining superior token efficiency.

Robot crowd navigation requires safe and efficient decision-making under dense, dynamic, and multimodal human--robot interactions. Existing reinforcement-learning methods typically output a single reactive action at each timestep, which limits their ability to represent diverse short-term avoidance strategies. We propose Planning Diffusion Policy Optimization (PDPO), an offline-to-online reinforcement-learning framework that uses a diffusion policy to generate short-horizon action chunks for crowd navigation. PDPO is first pretrained on collision-avoidance demonstrations and then fine-tuned online with PPO by treating the denoising process as an internal decision process. During execution, the policy generates a five-step action chunk and applies it in a receding-horizon manner. Furthermore, we observe an evaluation artifact in common crowd-navigation benchmarks: without explicit boundary constraints, learned agents may leave the valid domain and bypass dense crowds. To address this, we introduce a setting in which boundary violations are treated as collisions. Experiments show that PDPO obtains an improved success rate over strong baselines, and ablations demonstrate that action chunks are especially important for the modified bounded benchmark.

Correlation Clustering (CC) is a natural formulation of clustering in combinatorial optimization, which uses a graph representation of the input and does not require a pre-specified number of clusters. Given $n$ objects and a pairwise similarity function, the goal is to cluster the objects so that similar objects are put in the same cluster and dissimilar objects are put in different clusters. Despite its versatility, existing CC algorithms suffer from significant scalability issues and are inherently transductive: i.e., the algorithm must be executed from scratch for any new problem instance. In this work, we bridge this gap by leveraging Graph Neural Networks (GNNs) to solve Inductive Correlation Clustering, a novel generalization of the CC problem designed to handle unseen graph instances. By learning to exploit common structural patterns and node features during training, our framework generalizes to new graphs drawn from the same distribution with minimal computational overhead with respect to standard algorithms. We demonstrate the effectiveness and scalability of our approach through extensive experiments. Our framework not only excels in the inductive setting, e.g., lowering the inference time up to $5$ order of magnitude, while maintaining an approximation ratio within $~10\%$ of the best baseline solution, but also achieves competitive results on standard (transductive) CC benchmarks. Finally, we showcase a practical application of our framework as a learnable pooling mechanism for graph classification. Our results indicate that our method serves as an efficient pooling layer, enhancing the ability of GNNs to capture hierarchical structural information in networks.

Applications in fields such as movement ecology, Internet of Things or robotics share the need for systems that localize devices that are too small and power constrained to implement GNSS (Global Navigation Satellite Systems). Alternative low-power localization methods often rely on only measurements of RSS (Received Signal Strength) to infer the AoA (Angle of Arrival) of a transmitted radio frequency signal, but are limited by range and the power demand of the large number of RSS measurements required to infer an accurate AoA. In this paper we address these issues with a novel RSS-based method for tracking ultra lightweight and low-power moving receivers across a complex landscape, achieved by using a minimal number of RSS measurements from simple rotating high-gain transmitters with a range of 300m, and applying probabilistic modelling to infer their AoA. The receiver's movement path is then modelled using a Gaussian process and reconstructed using doubly stochastic variational inference, resulting in approximately 15m accuracy tracking of receivers weighing 38mg (including power source) over a scalable landscape range while consuming less than 180uW, increased to approximately 10m accuracy at less than 600uW by taking more RSS measurements. We anticipate that this method will support fields such as the behavioural study of flying insect species, which we demonstrate by applying the system to track Bombus terrestris nest return flights.

Object classification in event-based computer vision is a task that is attracting considerable research attention. Event-based object classification is a fundamental task in the fields of security and applied computer vision, which typically use synchronous frame-based cameras and computing pipelines for operation. This approach has several practical flaws. The size, weight and power consumption of the device could prohibit deployment at the extreme edge or in covert sensing environments. Besides this, there are security concerns inherent in cloud-based or other off-device computation approaches due to the requirement of sending and receiving potentially sensitive data. Furthermore, this transmission of data introduces latency and requires consistent connectivity to the cloud infrastructure to function. The use of Spiking Neural Networks (SNNs) hosted on neuromorphic devices attempts to solve several issues present in this conventional approach. Research into event-based object classification methods are hindered by the lack of high-quality vision datasets to use. To this end, the ANTShapes simulation tool has been previously proposed to create and label event-based vision datasets. In this paper, four novel datasets of varying difficulties are created using the tool and are benchmarked against existing spiking datasets commonly used for event-based vision research (N-MNIST, CIFAR10-DVS, DVSGesture and POKER-DVS). Classification is performed using a convolutional SNN. This work simultaneously provides four datasets with rich details for future experiments to use and validates the output of the ANTShapes dataset simulation tool as being suitable for its purpose.

Conversational AI systems, such as chatbots and virtual assistants, are becoming increasingly important to digital business processes. However, the established Business Process Model and Notation (BPMN) standard faces challenges when representing dynamic, context-sensitive interactions. This paper addresses this methodological and practical research gap by developing a standard-compliant BPMN extension (BPMN4CAI). Using Design Science Research methodology, this paper develops an approach that systematically extends existing BPMN elements and incorporates specialized components. The applicability and relevance of the BPMN4CAI framework are demonstrated and evaluated through a case study. The results show that the BPMN4CAI extension facilitates adaptive decision-making processes, robust context management, and transparent interactions for Conversational AI within business processes.

Register-transfer level (RTL) verification consumes a major part of modern system-on-chip (SoC) development effort. Yet, recent LLM-based verification-code generation often fails to produce runnable, design-consistent, and coverage-producing testbenches. We present AgentDV, a closed-loop agentic AI framework for automated RTL verification environment generation. AgentDV transforms single-shot LLM testbench generation into a tool-grounded verification pipeline by combining LLM-guided analysis, testbench construction, simulation, coverage measurement, and iterative refinement. The framework introduces three key ideas: 1) runnability filtering to reject invalid generated environments, 2) CSR-grounded checking to reduce hallucinated signals and incorrect expected behavior, and 3) coverage-guided iteration to regenerate tests based on measured verification gaps. We evaluate AgentDV using three LLMs on challenge DUTs and public OpenTitan peripheral and security IP blocks. From our analysis, we observed that direct single-shot prompting fails to produce a valid coverage-producing environment on benchmarks. AgentDV achieves 100% pass rate on four DUTs and an average of 80.9% pass rate on all DUTs using Claude Sonnet 4.6. Similarly, an average of 58.7% and 60.6% pass rate is achieved for Llama and Qwen models, respectively. In addition, an average of 74.5%, 69.1%, and 64.9% of line coverage and 88.4%, 82.3%, and 76.7% of branch coverage for the benchmarks under consideration for Claude Sonnet 4.6, Llama, and Qwen models, respectively.

The development of frontier models is commonly perceived to be the exclusive remit of a small number of heavily funded players, creating an information, economic and power asymmetry between developers and the diverse user base of modern AI. Recent public discourse acknowledges this concern, calling for SovereignAI (an organisation's capability to independently build, deploy and govern AI use), but offers little concrete advice on how this can be achieved in the short term under a diversity of funding settings. We argue that frontier performance is achievable by a wide range of institutions through Continual Learning on readily available open-weight models. Unlike limited approaches such as small-scale fine-tuning, prompt engineering, or tool-augmentation of a frozen model, our approach exploits a modern mid- & post-training stack while introducing safeguards that preserve both plasticity and stability at each stage, making the minimal number of high-impact interventions on the parameters. This yields gains comparable to those typically seen across multiple successive model generations, at compute and personnel budgets substantially lower than commonly thought, making ownership of large parts of the SovereignAI stack (model, tool infrastructure, values & data privacy) viable for far more actors. We demonstrate this with Thomson, a general-purpose frontier model trained with an enhanced focus on high-stakes professional work. Thomson performs competitively with recent frontier models across agentic tasks, safety, legal, tax & multilingualism, and large-scale Deep Research. Evaluations show a distinctive $π$-shaped pattern: distinct improvements across a wide range of capabilities, including those not explicitly targeted, while almost completely eliminating the forgetting problem common to narrow domain adaptation.

Tool-augmented LLM agents must rely on untrusted runtime Observations to complete open-ended tasks; however, when tool outputs no longer merely provide data but begin to specify concrete actions, they effectively become ``commands'' that can drive real-world side effects beyond user intent. We argue that this risk arises from conflating action induction with execution authorization. To address this distinction, we propose SARA, which treats action induction and execution authorization as distinct runtime roles and separates action provenance from execution authority. On the Observation side, a context-isolated Action Probe exposes action-inducing semantics and persistently records action-origin provenance across steps as a review signal; on the execution side, actual tool calls are authorized only against the user objective and audited evidence from authorized successful executions, while satisfying goal, execution-chain, and argument-level support. To preserve this separation across multi-step execution, SARA applies No-History-Promotion to prevent historical recurrence from laundering action origins into execution authority. Across AgentDojo and AgentDyn, SARA limits ASR to no more than \(0.63\%\) across four primary evaluation settings while maintaining competitive task utility, and consistently reduces ASR across additional Agent backbones.

In recent years, graph anomaly detection (GAD) based on frequency-domain filtering have achieved promising results. However, existing approaches still face three major challenges: First, they use static basic function to constructed graph filter which cannot effectively adapt to the frequency-domain distribution of graph data. Second, they fail to adequately consider the importance information of each attribute in the node feature vector, leading to the loss of fine-grained information. Third, they insufficiently utilize node labels for GAD. To address these issues, this paper proposes a novel graph anomaly detection method called JPGFN (Feature Transformation Enhanced Jacobi Polynomial Graph Filtering Network). First, a Feature Separation Transformation Network (FSTNN) is developed to better learn fine-grained node features by feature separation and applying nonlinear transformations to node features across different dimensions. Second, an adaptive Jacobi polynomial graph filtering module is constructed based on Jacobi polynomials to adaptively capture complex frequency-domain features of graph signals. Finally, a node label constraint module is developed to facilitate the use of node labels and enhance the performance of GAD. Experimental results on multiple real-world datasets demonstrate that the proposed method significantly outperforms mainstream approaches.

Despite their potential in standardized graph tasks, Large Language Models (LLMs) remain brittle to real-world shifts in node identifiers and task formulation. While deterministic graph tools are invariant to such shifts, extracting topological structures from noisy text is highly fragile for LLMs, which often overfit to surface patterns. Moreover, mitigating these parsing failures via multi-agent systems incurs prohibitive latency. To address this, we propose GRAIN, a single-agent framework optimized via reinforcement learning. GRAIN models reasoning as a semantic parsing and tool-execution pipeline, guided by a Structure Invariance Reward. By validating extracted intermediate graphs against ground-truth topologies, this reward forces the LLM to learn robust text-to-structure mappings rather than memorizing linguistic artifacts. We also introduce GRIT, a benchmark evaluating sensitivity to such linguistic shifts. GRAIN outperforms multi-agent baselines by 16.45\% in accuracy with approximately 24\% lower latency. Furthermore, it demonstrates superior structural generalization, halving the out-of-distribution (OOD) gap of SFT models (from 15.77\% to 7.80\%) and maintaining robustness on large-scale graphs beyond the training distribution.

Large language model agents are increasingly deployed as autonomous loops. Starting from one human goal, such a system repeatedly discovers work, plans, executes tool calls, verifies outcomes and persists state across many unattended iterations. The agent safeguards in wide use, however, are defined over a single trajectory, and their safety state is re-initialized when the next trajectory begins. We show that this is a failure of composition rather than an implementation detail. Our central result is a separation: against an attack whose evidence is fragmented across several iterations, every trajectory-scoped monitor has a true-positive rate equal to its false-positive rate, however expressive it is, because the evidence it would need never appears in the window it sees, whereas a monitor retaining cross-iteration state separates the two perfectly. We further show that the obvious repair of carrying a geometrically decaying risk score is insufficient, because the cooling-off period a patient adversary must wait is a constant that does not grow with the horizon $N$. We then present LoopHarness, which restores a persistent, non-decaying safety state at the loop level. Under mediated commits and an arbiter detection floor $δ_M$, it bounds the expected number of unauthorized irreversible actions by $B+m-1+m/δ_M$, a constant in $N$, of which the $B+m-1$ term is decided by a model-free rule and therefore survives a fully colluding verifier. We give a complete evaluation protocol on native Agent-SafetyBench tasks with paired clean and attacked episodes, an outer-state attack suite whose decisive evidence exists only across iterations, per-module ablations, and an adaptive white-box red team.

The recently envisioned goal-oriented communications paradigm requires machine learning inference to be performed directly on wirelessly transferred data. This paper presents an eXtremely Large (XL) Multiple-Input Multiple-Output (MIMO) system that operates as an Extreme Learning Machine (ELM) to execute Over-The-Air (OTA) binary classification. To reduce hardware complexity, the receiver is equipped with cascaded metasurfaces terminating in a single radio-frequency chain. A front metasurface layer applies a fixed nonlinear response to the incoming signal, acting as the ELM's activation function. Subsequent tunable linear metasurface layers physically approximate the trained network weights directly in the wave domain. Numerical evaluations across diverse datasets showcase that our XL MIMO architecture achieves classification accuracy comparable to idealized digital models, thereby proving the viability of low-complexity, wave-domain OTA learning.

Multimodal foundation models are increasingly used in speech-first assistants that must interpret spoken queries and produce visually grounded decisions. Yet it remains unclear whether semantically equivalent queries yield consistent judgments across modality (text vs. speech) and language (English vs. Arabic). We introduce a speech-augmented visually grounded contrastive triplet benchmark spanning 10,150 culturally grounded images from 18 MENA countries, where each image is paired with one supported statement and two plausible but unsupported alternatives. We define contrastive instability as the conditional rate at which a model fails to resolve all statements within a triplet, isolating fragmented reasoning from complete failure. Evaluating recent multimodal models under text and speech in English and Arabic, we find that modality and language shifts introduce substantial triplet-level inconsistencies that are not fully captured by aggregate accuracy, with speech amplifying partial failures. We make the benchmark publicly available to the community.

Long-context inference is bottlenecked by the memory footprint of the key-value (KV) cache, especially for small models under tight resource budgets. Existing KV cache eviction methods score tokens using the model's attention distribution or, in attention-free variants, each key's distance from a global reference point. Using a controlled leave-one-out probe, we find that attention magnitude is unrelated to a token's causal contribution to the answer (Spearman $ρ=-0.004$), challenging the premise behind dominant eviction methods. We introduce TwinKV, a training-free, attention-free redundancy signal that detects whether a token's key has a near-duplicate elsewhere in context. Rather than replacing existing policies, TwinKV acts as a composable repair pass: given a policy's fixed retained set, it identifies evicted tokens with no surviving duplicate (\emph{orphans}) and retained tokens whose information is duplicated elsewhere (\emph{redundant donors}), then swaps them while preserving the original budget and scoring rule. We compose TwinKV with four recent eviction policies across LongBench, LooGLE, RULER, and a short-context MMLU-Pro no-harm control at compression ratios ${0.3,0.5,0.7}$. On Qwen3-4B, TwinKV improves a majority of configurations for two policies, is near-even for a third, and helps only a minority for a fourth adaptive baseline already near a performance ceiling; gains across the three non-ceiling policies are smallest at the loosest ratio. On RULER with Llama-3.2-1B, however, that fourth policy improves in every evaluated cell because its Alone score leaves substantial room to improve. More broadly, Llama-3.2-1B shows a smaller average LongBench gain but a higher fraction of improved cells on LongBench and LooGLE than Qwen3-4B, plus a clean RULER win. We also identify few-shot classification exemplars as a task structure where TwinKV does not help on either model.

Internet memes are a pervasive form of multimodal online communication; however, such communication often involves users from diverse linguistic and cultural backgrounds. Therefore, adapting memes across cultures and languages is a central challenge for enabling mutual understanding in online communication. Unlike ordinary translation or standalone text rewriting, cross-cultural meme transcreation must jointly preserve communicative intent, adapt culture-dependent meaning for the target audience, and maintain coherence between text and image. In this work, we first provide an explicit task analysis of cross-cultural meme transcreation and identify three core challenges: culture-specific knowledge understanding, intent and tone preservation, and multimodal consistency. Based on this analysis, we propose a multi-agent framework with specialized agents that are coordinated to address these challenges through cultural adaptation, target text rewriting, revision, and conditional visual adjustment. The framework strengthens target text adaptation with coordinated feedback to handle difficult cases that require deeper cultural or visual intervention. We evaluate the framework on bidirectional Chinese-English meme transcreation using both human evaluation and LLM-as-a-Judge. Our method consistently outperforms all baselines across both evaluation settings. In human evaluation, it achieves the best performance on all four dimensions and delivers a 33.1% average improvement over the strongest baseline, while in LLM-as-a-Judge, it attains the highest Top-1 ranking rate (60% versus 26% for the second-best baseline). Further analysis indicates that each component contributes to the performance. Our error analysis suggests that the remaining bottlenecks lie in humor reconstruction and image-text alignment rather than simple cultural knowledge gaps, pointing to future work on humor transfer.

Modern software engineering establishes software supply chains and relies on tools and libraries to improve productivity. However, reusing external software in a project presents a security risk when the source of the component is unknown or the consistency of a component cannot be verified. Reproducible builds present a mitigation strategy, as they can confirm the origin and consistency of reused components. A large reproducibility community has formed for Debian, but the reproducibility of the Maven ecosystem, the backbone of the Java supply chain, remains understudied in comparison. Reproducible Central is an initiative that curates a list of reproducible Maven libraries, but the list is limited and challenging to maintain due to manual efforts. Our research aims to support these efforts in the Maven ecosystem through automation. We investigate the feasibility of automatically finding the source code of a library from its Maven release and recovering information about the original release environment. Our tool, AROMA+, can obtain this critical information from the artifact and the source repository through several heuristics and we use the results for reproduction attempts of packages on Maven Central. Overall, our approach achieves an accuracy of up to 99.8% when compared field-by-field to the existing manual approach. In some instances, we even detected flaws in the manually maintained list, such as broken repository links. We reveal that automatic reproducibility is feasible for 32% of the packages on Maven Central using AROMA+, and 12% of these packages are fully reproducible. We demonstrate our ability to successfully reproduce new packages and have contributed some of them to the Reproducible Central repository. Additionally, we highlight actionable insights, outline future work in this area, and make our dataset and tools publicly available.

Reliable prediction of time-varying channel state information (CSI) is essential for efficient wireless communication. Each CSI frame is a matrix-valued representation of the wireless channel response, and a sequence of CSI frames forms a temporal channel trajectory. Modern deep learning-based CSI predictors, however, often provide only point predictions and lack calibrated uncertainty estimates. This limitation is particularly problematic in multi-step CSI prediction, where the target is a sequence of future CSI matrices, and downstream decisions such as beamforming or scheduling may fail if any part of the predicted trajectory is unreliable. We propose trajectory-adaptive calibration and error profiling with conformal risk control (TRACE-CRC), a method for trajectory-aware uncertainty quantification in multi-step CSI prediction. TRACE-CRC constructs Frobenius-norm uncertainty balls around predicted CSI matrices and controls the risk that at least one future frame is uncovered. Instead of calibrating each future step independently, TRACE-CRC combines future-step-dependent error profiling, trajectory difficulty stratification, and learn-then-test (LTT) risk control. Empirically, TRACE-CRC achieves reliable trajectory-level coverage with substantially smaller uncertainty balls than conservative multi-step corrections, while avoiding the trajectory undercoverage of compact stepwise and adaptive conformal baselines.

Directed graph learning naturally leads to trainable nonsymmetric propagation operators with distinct right and left spectral structures. Building on the two-sided cone Rayleigh framework for generalized pencils \[ B_θ-λG, \] we develop a learning-oriented methodology for spectral certification, sensitivity analysis, and control without requiring symmetry, nonnegativity, or cone preservation. In the positive-orthant setting, computable lower and upper cone bounds provide an a posteriori enclosure of a distinguished cone level, while smooth soft-min/max surrogates preserve rigorous one-sided bounds with explicit approximation errors and remain differentiable with respect to the trainable parameters. For a simple interior level, the right and left modes satisfy \[ Dλ_C(B)[H]=v_C^T H u_C, \] yielding first-order optimal graph-supported interventions under prescribed perturbation budgets and motivating adaptive spectral control. Numerical experiments demonstrate the applicability of the approach beyond cone-preserving operators and in directed learning settings. Signed nonsymmetric perturbations reveal a transition from interior eigenpairs to boundary complementary quasi-pairs, including non-spectral cone levels, while controlled experiments show that symmetrization can remove predictive information carried solely by edge direction. On the directed Cora citation network, adaptive recomputation of the right--left sensitivity reduces the distinguished spectral level by approximately $21.5\%$ under a cumulative edge-weight reduction budget of $0.5\%$, with no observed change in test accuracy for the trained model and data split considered.

Aesthetics are an important part of the symbolism of artistic works. Although subjective, humans categorize art based on the emotion evoked regardless of modality. What remains under-explored is how AI models form their own aesthetic categorization of human-produced media without explicit labels or cross-modal supervision. We present a self-supervised framework that projects four modalities (text, audio, image and video) into a shared 256-dimensional embedding space and applies iterative clustering to discover aesthetic structure. We discuss the divergence between AI-generated cluster assignments and human affective register labels on a weakly supervised multimodal dataset. This work has applications in understanding how AI structures cross-modal similarity, organizing heterogeneous media collections for Retrieval-Augmented Generation (RAG), and automated data labeling.

Cross-lingual alignment in multilingual language models is typically attributed to joint training: shared parameters, mixed-language batches, or explicit alignment objectives. We ask whether monolingual models trained on non-parallel data learn alignable representations without joint training. By testing on strictly monolingual language models, such as the Goldfish model families and independently developed models from different research labs, we find three results. Correlation: these models develop alignable representational geometry across layers, with alignment strengthening as data scale, model scale, or linguistic proximity increases. Construction: a single Procrustes rotation fit on parallel sentences maps hidden states between models. Causation: the same rotation transfers functional content; patching a rotated English residual into a German model on a factual cloze flips the prediction to the donor's capital in most cases. We confirm that cross-lingual alignment can emerge from the structure of language and the information it carries rather than from joint training, and this points to practical future directions including model stitching, merging, and modular multilingual systems built from monolingual components.

Motivated by theoretical problems in deep learning, we conjecture that post-composing a fixed number of pairwise distinct nonconstant polynomials with a generic polynomial of sufficiently large degree yields linearly independent polynomials. This generalizes Newman--Slater's theorem on powers of polynomials. We establish several cases of this conjecture and its origin-passing variant: We prove the result for two polynomials, and for an arbitrary number of polynomials when their degrees are bounded. Furthermore, we show how the conjecture implies a complete understanding of the identifiability (i.e., parameter symmetries) of deep fully connected neural network architectures with generic polynomial activation functions. In particular, for network architectures with layer-specific activations of increasing degree, our established versions of the conjecture fully characterize the set of parameters yielding the same end-to-end network function. As a special case, we fully resolve the identifiability of shallow polynomial networks.

Reliable medical conversational AI requires authentic expert--patient interaction data, yet such datasets remain scarce, especially for low-resource languages such as Bengali. We present DocTalkBN, a large-scale multimodal dataset of real-world expert telemedicine conversations in Bengali, collected from nationally broadcast telemedicine programs featuring board-certified physicians. DocTalkBN contains 557.63 hours of paired audio and text, 1,515 multi-turn patient calls, 10,274 host--doctor question--answer exchanges, totaling 1.7M tokens, spanning 26 medical specialties. Unlike prior resources derived from medical forums, written health content, or synthetic data, our dataset preserves the spontaneity, contextual richness, and spoken characteristics of authentic medical interactions in a low-resource setting. To support benchmark-driven research, we further construct three downstream tasks from the corpus, medical triage classification, advice safety evaluation, and medical named entity recognition, and benchmark a diverse set of large language models and encoder-based baselines. Our results show that DocTalkBN is a practically useful resource, particularly for clinically grounded reasoning tasks. We release this resource to facilitate future research on reliable medical NLP and safer, more culturally grounded healthcare systems for low-resource languages. Our source codes and dataset are publicly available at https://anonymous.4open.science/r/doctalk.

Federated Learning (FL) enables privacy-aware distributed training, yet gradient updates remain exploitable: Man-in-the-Middle (MitM) interception exposes updates in transit, while model poisoning corrupts global convergence. We first introduce GASHE (Gradient-Aware Selective Homomorphic Encryption), a novel selective encryption strategy that dynamically identifies and encrypts only the gradient components exceeding a DP-calibrated sensitivity threshold, rather than encrypting all parameters uniformly as in static layer-based or full-parameter CKKS schemes. Building on GASHE, we introduce SecureDrive-FL, a federated driver monitoring framework that couples DP-SGD with GASHE to create the first closed-loop DP+HE privacy pipeline: DP-SGD calibration parameters directly derive the GASHE encryption mask, unifying training-time privacy and communication-time confidentiality. Evaluated on a ten-class distracted driver classification task under non-IID federated splits, SecureDrive-FL matches DP-SGD alone's poisoning resistance (73.6% vs. 74.0% accuracy, 3.9% Attack Success Rate for both) while additionally withstanding MitM interception, where DP-SGD alone collapses to near-random accuracy (78.2% vs. 10.4%), all under only approx. 8--10% additional runtime overhead relative to DP-SGD alone---under DP-SGD noise injection with per-round privacy parameter epsilon_0=4.

Large Language Models (LLMs) operate in hospitals, courtrooms, banks, and public service desks, where fluent, confident outputs are treated as authoritative even when ungrounded or incorrect. When such an output contributes to harm, who is answerable, and through what mechanisms can responsibility be traced, explained, and acted upon? Following PRISMA guidance, five databases were searched from January 2022 to March 2026 against four review questions; of 4,512 records identified, 122 primary studies were included, together with 12 regulatory and standards documents analysed as primary sources. The review consolidates a sociotechnical account of accountability as an actor-forum relation resolved into five dimensions, and synthesises mechanisms across four families: technical controls, human oversight, organisational governance, and documentation and traceability, each with a maturity assessment. The corpus is read through a four-layer classification device spanning provenance, application logic, human oversight, and governance and redress, cross-cut by traceability, role clarity, and continuous monitoring. Both are mapped onto the EU AI Act, whose high-risk obligations have applied since 2 August 2026, the NIST AI RMF with its Generative AI Profile, ISO/IEC 42001, and sectoral guidance in healthcare, consumer finance, education, and the public sector. Four persistent gaps emerge: under-specification of human oversight, absence of shared accountability metrics, disciplinary disconnection, and limited empirical evaluation, alongside five structural tensions that no surveyed instrument resolves. The review closes by consolidating the classification device into an integrated accountability architecture, LAAF, with cybersecurity aligned to the OWASP LLM Top 10 (2025); it is a synthesis of the surveyed evidence rather than a validated artefact.

Ontology learning from text remains challenging despite significant progress in Large Language Models (LLMs), which can hallucinate domain terms, produce inconsistent formats, and favor hierarchical over associative relations. In the LLMs4OL 2026 Challenge, we address both the End-to-End Flagship Task (Task A) and Ontology Extension Reuse Task (Task B) using an offline retrieval-augmented few-shot prompting pipeline. Our system employs Qwen2.5-14B-Instruct with all-MiniLM-L6-v2 for demonstration retrieval, selecting the top-5 examples for Task A and top-2 for Task B. A left-truncated context-windowing strategy preserves task instructions within long prompts. For Task B, generated triples undergo deterministic vocabulary-constrained filtering, retaining triples when at least one endpoint belongs to the sample's closed term/type vocabulary and removing duplicates of the initial ontology. The approach achieves Semantic Graph Similarity of 0.8692, Term-Typing F1 of 0.9200, and Taxonomy Discovery F1 of 0.8540 on Task B, while Task A achieves 0.7416 Semantic Graph Similarity. However, no non-taxonomic relations are extracted, highlighting limitations of closed, taxonomy-oriented relation vocabularies.

Reproducibility in machine-learning research depends on experimental choices such as random seeds, dependency versions, data partitioning, and evaluation configuration. Existing repository validation workflows may execute successfully without detecting changes to such choices. We study this problem using MLReproMutate, research software that applies controlled, reproducibility-relevant mutations to ML research repositories and evaluates them against validation workflows already present in those repositories. We conducted an outcome-blind empirical study of 39 frozen repository-operator cases using four mutation classes: random seed, dependency pin, data split, and cross-validation fold count. Repository revisions, mutation candidates, and validation workflows were fixed before mutation outcomes were observed. Primary execution yielded outcomes for 13 of 39 cases; a bounded restoration procedure increased the combined evaluable set to 24. After excluding one confirmed-equivalent mutation, 23 confirmed non-equivalent mutations remained. The selected validation workflows detected 2 of these 23 mutations, corresponding to an observed detection proportion of 8.7%. These results do not imply that the corresponding repositories are irreproducible. Rather, they show that, in this sample, existing validation workflows often did not detect the particular controlled reproducibility-relevant changes introduced by the study. The findings motivate reproducibility-oriented mutation testing as a complementary way to assess whether research-software safeguards constrain experimentally important choices.

Large language models have shown strong potential for software engineering tasks, particularly software testing. Model-based testing (MBT) is a software testing technique. To address the broad scalability challenge for industrial adoption of MBTs, our paper presents an empirical evaluation of Large Language Models (LLMs) for automated model-based test generation, compared with a state-of-the-art model-based testing tool (GraphWalker) and its built-in algorithms (random and quick random for edge and vertex coverage settings). Our evaluation indicates strong potential to optimize and shorten test paths and step sizes using the recent five state-of-the-art LLMs (GPT-5.1, GPT-5.2, Claude Opus 4.5, Claude Sonnet 4.5, and Gemini 2.5 Pro) against four GraphWalker models (two web applications (Parabank and Testinium) and two hardware applications (TLC and RISC-V) ) of escalating complexity.

Bug localization is a labor-intensive task, particularly in large software systems. When abnormal behavior occurs, developers must perform repetitive and time-consuming steps to identify faulty files. Previous studies have mainly focused on single-objective localization methods, many of which are limited to specific programming languages. In addition, relying solely on lexical similarity between source code and bug reports is often insufficient due to the natural language nature of bug descriptions. In this study, we propose a class-level automated multi-objective search-based system to identify and rank potentially buggy classes from bug reports. The main objective is to maximize similarity while minimizing the number of suggested faulty files. The evolutionary optimization algorithm SPEA-2 was applied to six open-source Java projects comprising more than 22,000 bug reports. The proposed approach was evaluated against two widely used algorithms, NSGA-II and MOEA/D. Results indicate that SPEA-2 achieved higher precision and recall than both multi-objective and single-objective baseline methods. The proposed recommender system successfully identified buggy classes or files for 88.5\% of bug reports within the top 10 recommendations and 94\% within the top 20. The effectiveness of the model was further validated on an industrial Android project written in Kotlin, demonstrating its adaptability across programming languages.

While most phenotyping platforms rely primarily on image-based measurements, advanced plant characterization requires the integration of active physiological sensing modali- ties such as chlorophyll fluorescence. We present an autonomous robotic platform designed to perform targeted fluorescence measurements on plant leaves. The system combines 3D plant reconstruction, geometric analysis, and motion planning to localize suitable measurement points and generate collision-free trajectories for a robotic manipulator. A dense 3D model of the plant is reconstructed from multi-view data and used to extract candidate leaf surfaces based on orientation, accessibility, and sensing constraints. These targets are then integrated into a task-level planning framework that guides the end-effector to precise contact or near-contact configurations required for point-based fluorescence acquisition. The platform enables automated, repeatable, and spatially resolved physiological measurements that go beyond passive imaging. By tightly coupling perception, geometric reasoning, and manipulation, the proposed system provides a robotics-driven approach to high-resolution plant phenotyping and opens new directions for autonomous agricultural inspection and plant-aware manipulation.

Enterprise AI deployment is a coordination problem across business units, application and AI teams, testing, platform engineering, infrastructure, security, operations, and data governance. Use-case benchmarks show whether one agent completes one task, but not how changing capabilities, models, runtime mechanisms, capacity, and enterprise data should be owned, changed, admitted, or evidenced together. We present four responsibility objects as shared organizational contracts: Skill (reusable, versioned capability and workflow asset), Harness (runtime compiler and governor), Scaffold (execution/control boundary and NFR owner), and a stack-external data substrate under independent CIO-governed semantics and telemetry. The runtime core is A = <S, H, X>, with the data substrate outside that stack. The central contribution is one bounded, falsifiable hypothesis, P1 (cost-aware capability-capacity separability): within a declared operating region, changing activated capability preserves the capacity-response interaction within a preregistered equivalence margin, while changing compatible Scaffold capacity preserves capability semantics up to a non-inferiority margin, and the required controls stay within a declared enforcement budget. Six design conditions become measured obligations whose coverage, violations, uncertainty, cost, and exclusions determine whether P1 is decidable. We propose a cluster-period randomized crossover experiment (balanced order, reset/washout, repeated seeds and failure regimes, cluster-aware uncertainty) with a four-state verdict: supported, falsified, conditional-engineering, or inconclusive. This paper contributes a contract-bounded runtime architecture, a source-preserving data substrate, and a falsifiable measurement protocol. It reports no completed implementation, experiment, dataset, or measured result.

Many scientific and engineering applications require estimating unknown parameters from experimentally observable data -- an inverse problem that is inherently challenging due to nonlinearity, noise, and ill-posedness. In this paper, we propose an active diffusion-based inverse problem solver. A DM is trained to learn the mapping between the parameter space and the observable space. By iteratively detecting and correcting model misspecification through posterior uncertainty, the method discovers and learns the correct region of parameter space, even when initial training bounds exclude the true parameters. This provides a principled, Bayesian justification for adaptive domain augmentation and ensures robust inference for inverse problems under incomplete prior knowledge. We demonstrate the effectiveness of our inverse solver for a toy inverse problem with infinite solutions, and for the parameterization of the quantum correlation functions to event observables in a Quantum Chromodynamics analysis of nucleon structure.

Extracting transverse momentum dependent parton distribution functions (TMD PDFs) from semi-inclusive deep inelastic scattering (SIDIS) data is a central goal of the nucleon structure program at Jefferson Lab and the future Electron-Ion Collider. Traditional extraction methods rely on parameterized functional forms and iterative fitting, which can limit the flexibility of the resulting distributions and make uncertainty quantification cumbersome. We present a conditional diffusion model that learns to map raw SIDIS event kinematics directly to TMD PDFs, bypassing explicit functional assumptions. Evaluated on simulated SIDIS data at CLAS12 kinematics, the model recovers the underlying TMD with informative uncertainties that narrow steadily with increasing event statistics, and produces reliable estimates even with as few as 1,000 conditioning events, a statistics-limited regime directly relevant to ongoing and planned experiments.

Algorithmic trading now represents a market exceeding $20 billion, where even marginal gains in signal robustness can translate into economically significant returns. Existing evaluations of equity prediction models do not explicitly target regime robustness during hyperparameter selection. Five model classes are trained on daily observations from approximately 300 large-cap US equities over eleven years, with Bayesian optimisation configured to target trading performance across three statistically different market regimes. Regime-robust hyperparameter selection is associated with out-of-sample generalisation, as signal precision remains above the random baseline across all four quarters of the test period, and portfolio performance slowly degrades under simulated input noise before collapsing beyond a defined threshold. No individual tabular deep learning architecture outperforms gradient-boosted trees, but combining XGBoost and TabNet using rank aggregation produces a Hybrid ensemble with an annualised return of 51.26%, a Sharpe ratio of 2.44, and a statistically significant CAPM alpha of 0.423 (p = 0.011). A near-zero beta indicates this outperformance is driven by stock selection, not market exposure. Alternative data plays a secondary role once technical and fundamental features are accounted for, as well as contributing more strongly on the short side than the long, and varies by model class. An interactive application makes these results explorable in real time, with live data integration the remaining step toward practical deployment.

A core question in decision-making for agents is whether the relative priorities of competing lower-level objectives can be determined by emotional preferences autonomously generated by higher-level goals, rather than being externally prespecified. Under changing external environments and evolving internal states, emotions play an important functional role in regulating the relative priorities of competing goals. Inspired by the goal-directed theory of emotion, this paper studies how such preference regulation can be computationally realized through reinforcement learning. We first propose a conception of emergent emotional preference: a high-level goal autonomously induces state-dependent preferences over competing lower-level objectives. This conception is built upon a framework consisting of a multi-objective reinforcement learning inner controller and an outer preference generator. The inner controller provides a repertoire of preference-conditioned goal-directed behaviors, while the outer preference generator learns a mapping from the current state to objective preferences through reinforcement learning on a high-level goal. We operationalize emotional preference as a state-dependent regulation of relative goal priorities that emerges through optimization. Furthermore, we characterize the policy space induced by preference regulation and derive an upper bound on the optimality gap in terms of the representation error of the inner behavioral repertoire. We show that the gap vanishes when the optimal policy can be represented by the available preference-conditioned policies. Experiments in self-constructed multi-objective exploration environments show that the learned preference function exhibits contextual priority switching, graded trade-offs, and temporal persistence, and outperforms the evaluated fixed-preference and handcrafted-preference strategies.

To mitigate catastrophic forgetting in downstream continual learning (CL) for large language models (LLMs), existing methods typically constrain parameter updates or introduce task-specific adaptation modules. However, these methods often rely on explicit task boundaries during training, limiting their applicability to realistic task-free scenarios. In this paper, we propose a \textbf{Fi}sher-guided \textbf{uni}fied (\textbf{FiUni}) framework for batch-level task detection and parameter-efficient continual adaptation. FiUni is motivated by a key observation about the Fisher information matrix (FIM) of pre-trained models: the orthogonality among the principal subspaces of its Kronecker-Factored Approximate Curvature (K-FAC) approximation, estimated from a small number of downstream task samples, can reflect the similarity between different tasks. Based on this observation, FiUni constructs FIM-derived frozen subspaces to guide low-rank adaptation (LoRA), while matching the Fisher principal subspace of each incoming batch window with historical subspaces. This enables FiUni to adaptively determine whether to reuse existing knowledge, expand a related subspace, or create a new subspace, dynamically balancing knowledge sharing and task isolation. Experiments show that FiUni can effectively infer latent batch-level task affiliations and achieve competitive performance against advanced task-aware CL methods with fewer trainable parameters.

Privacy-Enhancing Technologies (PETs) in computer vision often rely on noise or image perturbations to protect visual data while securely processing it, creating a trade-off between task performance and protection. This trade-off is commonly evaluated using image classification, which primarily captures semantic separability and remains robust despite significant geometric, spatial layout or local boundary alterations. As a result, it is too simplistic as a proxy for generic vision tasks. Exhaustive downstream-task evaluation, however, is computationally expensive because models must often be trained for each PET transformation and parameter setting. We therefore propose a compute-aware multi-task protocol for evaluating PETs in model training. It combines lightweight proxy tasks that target complementary aspects of visual structure while remaining simple and fast to compute. Across irreversible privacy transformations, key-based block primitives, and learnable image encryption schemes, we demonstrate that PETs with similar classification accuracy can differ substantially on other tasks. The outcomes highlight the need for PET evaluation protocols that move beyond classification-only reporting.

Reliable assessment of causal research designs in the social sciences is critical for evidence-based policy-making, yet has so far relied entirely on manual expert analysis. We introduce Automated Research Design Tracking and Assessment (ARDTrA), a task that involves detecting the research design used in a paper and assessing the quality of its application. We create an expert-annotated dataset of papers covering six families of counterfactual research designs and evaluate the task using a multi-turn RAG-based conversational pipeline. Across four retrieval strategies, four LLMs and six embedding models, we find that passage length is the main driver of performance, explaining 52-66% of the variance. A per-research-design analysis also shows that human and machine difficulty do not align: the designs that prove hardest for the system are not those on which expert annotators disagree most, pointing to two independent sources of task difficulty.

Silent speech recognition (SSR) provides an alternative communication pathway in the absence of audible speech. However, conventional approaches are limited by the need for constant facial attachment, privacy concerns, and unstable signal acquisition. Here, we propose a soft, active electromyography (EMG) interface that enables word-level SSR using machine learning. Worn on the hand, the device uses a fingertip electrode that can be positioned near the lips to acquire EMG signals only when needed. The interface integrates liquid metal (LM) interconnects, transparent flexible printed circuit (FPC) electrodes, and elastomer encapsulation to ensure high mechanical stability during finger motion. A deep neural network trained on these stable signals achieved a mean accuracy of 97.2 $\pm$ 1.3% across three subjects in classifying a 30-word vocabulary, demonstrating robust linguistic discrimination. Furthermore, real-time drone control validates the practicality of this approach in noisy and privacy-sensitive environments where conventional voice recognition fails. This study highlights the potential of soft, wearable EMG systems as secure and intuitive human-machine interfaces.

Reinforcement Learning with Verifiable Rewards (RLVR) and other RL-style post-training paradigms have been used for aligning large language models (LLMs) with reasoning standards. The resulting recent Reasoning Language Models (RLMs) such as DeepSeek-R1, o3, and Kimi k1.5 show that such RL-style post-training ("RL-for-LLMs") can substantially improve chain-of-thought reasoning, long-horizon planning, and self-correction. However, the computational footprint of these systems is massive: state-of-the-art RLM training requires millions of GPU-hours and tightly coupled multi-model pipelines that stress modern hardware far beyond classical supervised LLM training. This makes RLM training as much a parallel and distributed systems problem as an algorithmic one. In this work, to facilitate developing RLMs that are simultaneously high-performance, scalable, and cost-effective, we first systematize the RL-for-LLM paradigm and provide a compute-centric analysis of prominent post-training algorithmic frameworks: Proximal Policy Optimization (PPO), Group Relative Policy Optimization (GRPO), as well as their variants. Second, we develop a taxonomy of intra- and inter-model parallelism strategies for RL-for-LLMs, covering both traditional techniques (data, tensor, pipeline, sequence, context, and expert parallelism) as well as novel forms of parallelism and optimization techniques for multi-model RLM training, for example disaggregated placement, stage fusion, hybrid parallelism, and asynchronous execution. We harness the work-depth model of parallel computing to make our taxonomy and its insights rigorous and portable. Finally, we analyze existing RLM frameworks and we distill practical guidelines and outline open research directions for building scalable, fast, and cost-effective RLMs.

Multimodal representation learning has been shifting from traditional two-tower architectures to large language model (LLM)-based embedders due to their strong instruction-following capabilities. Despite this progress, existing approaches primarily focus on language and image modalities, which also remain the dominant modalities for user-conditioned interactions in current embedders. In this paper, we propose the first Omni-Interactive Universal Embedder (OmniUE), which not only learns a unified embedding space across text, video, and audio by leveraging intermediate-layer representations from dedicated learnable tokens, but also supports omni-interactive querying, enabling users to provide inputs in the form of text, visual regions of interest, and audio spans. Within OmniUE, visual and audio segmenters process diverse user interactions and integrate them with an omni-LLM to produce user-conditioned any-to-any embeddings via context aggregation. To evaluate OmniUE's omni-interactive capabilities, we introduce OmniCHOIR, benchmarking models for omni-interactive compositional audio retrieval based on the given text, video, and audio as well as unimodal or multimodal interaction prompts. OmniUE consistently surpasses state-of-the-art baselines across diverse modalities, with average improvements of 10.5% on textual-interactive video benchmarks (MMEB-v2-video), 1.1% on audio tasks (MAEB), 83.7% on visual-interactive benchmarks (SCaR), and 24.1% on our omni-interactive OmniCHOIR benchmark. We believe that jointly advancing omni-modal representation learning and omni-interactive querying paves the way toward universal embedders.

Accurately forecasting the movement of people in complex scenes requires reasoning over the past and present state of the entire environment. In this context, effectively incorporating object information and social interactions into a unified framework remains particularly challenging. To address this, we propose Object-Conditioned Social Diffusion (OCSD), a conditional diffusion model that integrates motion history, multi-person interactions, and object cues into a single framework. OCSD uses an object-conditioning mechanism that modulates denoising at every timestep, enabling fine-grained human-object reasoning, and a social encoder that models the interactions between all humans in the scene. As a result, our model naturally handles varying group sizes, complex social interactions, and supports sampling multiple plausible futures. Extensive experiments show that OCSD achieves state-of-the-art results on the Humans in Kitchens (HiK) and HOI-M3 benchmarks. It reduces the two-second path error by 121.5 mm (31.3%) on HiK and 130.5 mm (33.2%) on HOI-M3 compared to prior work, and produces more realistic long-term forecasts.

Although batch prompting makes large language model inference more efficient by processing multiple instances simultaneously, it suffers from unpredictable downstream task performance. We propose cascaded batch prompting, a two-stage approach designed to resolve the unpredictability of conventional batch prompting by disentangling complex reasoning from symbol grounding. Experiments on multiple-choice question answering and natural language inference demonstrate that the proposed method outperforms the standard single prompting baseline while achieving a speedup proportional to batch size, establishing a new state of the art on the Pareto frontier.

Large Language Models (LLMs) can be trained to perform chain-of-thoughts reasoning in order to improve the reliability of their responses. In this work, we investigate how explicit reasoning can be leveraged for LLM-Based Machine Translation (MT) with in-context samples. We introduce a novel fragment-based reasoning framework in which the model first extracts parallel source-target fragments from retrieved similar exemplars, and uses these fragments as intermediate reasoning traces to produce the final translation. To train our model, we distill silver fragments and drafts from a large teacher model. Our experiments with the Qwen3 model family, over 6 languages, including up to 5 domains per language, demonstrate that fragment-based MT significantly outperforms alternative methods like standard k-shot or basic drafting.

Korean constituency parsing raises a representational challenge because the terminal units of a phrase-structure tree do not straightforwardly correspond to simple surface words. Korean eojeols are morphologically complex spacing units, and existing constituency resources differ in how they represent eojeol-internal morphology and non-overt elements. This paper compares three constituency parsing representations derived from the Penn Korean Treebank: Morpheme+XPOS, Eojeol+XPOS, and Eojeol+UPOS. We construct these representations by removing null elements, aligning Penn Korean phrase structure with overt eojeol tokens, preserving Penn Korean phrase labels where possible, and varying the terminal and preterminal layers. We then evaluate canonical non-binary transition-based constituency parsers in top-down, in-order, and bottom-up orders under a shared modeling and evaluation setup. All experiments use gold terminal segmentation and gold preterminal labels and therefore evaluate constituency parsing conditioned on gold morphosyntactic annotation. Eojeol terminals yield shorter transition sequences, but Eojeol+UPOS parsing substantially underperforms the morphologically richer conditions. Eojeol+XPOS narrows this gap, while Morpheme+XPOS gives the strongest results even after its predictions are projected to the eojeol terminal domain. Under these gold-annotation conditions, the results show that fine-grained morphological and XPOS representations provide valuable evidence for the evaluated parsers. This empirical finding concerns the information available for parsing and does not by itself determine the linguistically preferable terminal domain. Independently, linguistic and resource-design considerations motivate eojeol as a stable and interpretable surface domain for phrase-structure annotation, with morpheme-level and XPOS information retained as aligned morphosyntactic evidence.

Direct Preference Optimization (DPO) is a widely used objective for aligning language models from preference data, with the coefficient $β$ commonly interpreted as controlling the KL constraint to a reference policy. We show that $β$ entangles two distinct roles: it governs the effective inverse preference-noise scale and simultaneously rescales the optimization dynamics, coupling this scale with the effective step size. As a consequence, at a fixed learning rate the achieved policy deviation is non-monotone in $β$: it vanishes in a dead zone at small $β$, reaches a peak at an intermediate value, and decreases again for larger $β$. Moreover, standard DPO loss values are not comparable across $β$: runs with nearly identical loss curves can differ several-fold in KL divergence from the reference model. This entanglement obscures the role of $β$, increases sensitivity to hyperparameter choices, and complicates learning-rate scheduling. We propose a centered-softplus reformulation that is argmin-equivalent to DPO for $β>0$, while making the inverse preference-noise-scale and learning-rate effects explicit and independently tunable. The normalized centered-softplus objective also admits a continuous $β\to0$ endpoint that reduces to a linear preference-margin objective.

Measuring alignment between documents and structured reference frameworks requires identifying conceptual evidence distributed throughout the text and reporting it through measures that are quantitative, interpretable, and traceable. Many commonly used retrieval and classification approaches return either pairwise similarity scores or one or more class labels, whereas fewer methods provide concept-level scores that are directly traceable to the terminological evidence supporting them. We present \emph{Intelligent Target Locator} (ITL), a domain-agnostic and language-portable methodology that estimates the affinity between the textual units of a target document and the concepts defined in a \emph{Structured Reference Document} ($SRD$). From the $SRD$, ITL induces concept-specific terminological profiles built from independent terms, bigrams, trigrams, and co-occurrences. Each term is assigned an importance weight that combines concept membership, term-type specificity and inter-concept discriminability. The output is a textual-unit--concept affinity matrix that can be aggregated at different levels of granularity. We conduct an internal consistency assessment using the 17 Sustainable Development Goals (SDGs), evaluating each official goal statement against the $SRD$ induced from the same set of descriptors. Every statement reached its highest affinity with the corresponding concept, and the mean affinity across the remaining concepts stayed marginal relative to the mean reference affinity. This separation indicates that ITL distinguishes the conceptual profiles of the framework. ITL thus offers a general basis for quantifying document alignment with structured frameworks while keeping each result traceable to the terminological evidence that supports it.

Training molecular machine-learning models on ephemeral or memory-constrained accelerator instances can require repeatedly retrieving preprocessed molecular graphs from remote storage. FoldPipe is a lightweight Python orchestration layer for already-sharded PyTorch and PyTorch Geometric data. It retrieves one shard ahead in a background thread while the consumer trains on the current shard, keeping the number of live shard payloads bounded with respect to total dataset size. Asynchronous prefetch and bounded buffering are established systems techniques rather than novel scheduling algorithms. FoldPipe's contribution is a small integration targeted at native .pt molecular shards together with a source-pinned empirical characterization of its operating regime. We evaluate a SchNet energy-and-force workload on MD17 aspirin using 20 paired, order-alternating benchmark passes on a Tesla T4. Each pass processes five pinned shards containing 25,000 structures. FoldPipe records 16.33 s mean I/O-compute overlap, compared with zero by construction for the sequential bounded baseline. Mean pass time is 76.78 s for FoldPipe and 83.37 s for the baseline. However, the geometric mean paired speedup is $1.059\times$ with a 95% bootstrap interval from $0.878\times$ to $1.288\times$. The experiment therefore verifies the overlap mechanism but is inconclusive about a reliable wall-clock speed advantage under the observed public-network variability.

LLM-based agents are increasingly proposed for network fault diagnosis, but existing benchmarks evaluate them only on accurate tickets and always assume a fault is present, conditions rarely met in practice. We present FaulT-Bench, a benchmark of 200 troubleshooting scenarios across eight network topologies, five reimplemented from public practitioner labs, spanning genuine faults, false fault reports, incorrect device attribution, and incorrect root-cause claims. To isolate how ticket wording affects diagnosis, we further rewrite 72 false-premise tickets into five reporter personas that vary reporter confidence and verifiable detail one factor at a time, holding the network state fixed. Our automated harness deploys each scenario in Kathará, lets agents interact through the NIKA tool interface, and scores free-text diagnoses with an LLM judge across outcome, fix, and reasoning quality. Evaluating SADE, ReAct, and Claude Code, we find all three are near-saturated on accurate tickets and robust to misdirection, yet degrade sharply when the network is healthy and the ticket is wrong, probing until a benign condition can be promoted to a root cause rather than concluding nothing is wrong. Persona rewrites show that how a ticket is written matters more than what it claims: a confidently wrong report is handled about as well as an accurate one, while a vague, underspecified report degrades performance sharply. The three agents also fail differently, from constant over-diagnosis to unanswered runs, at very different cost. These results position FaulT-Bench as a benchmark for developing agentic systems that can reason reliably over the noisy, unreliable tickets of real-world network troubleshooting.

Hidden coordinates are not uniquely determined by a language model's input--output function, so representation-derived measurements should be invariant to function-preserving changes of basis. This study shows that column-permutation parallel analysis violates function-preserving reparameterization invariance because its reference distribution and selected component count can change while the model function and observed covariance spectrum remain fixed. More generally, a data-internal reference procedure cannot simultaneously preserve every coordinate marginal, remain orthogonally equivariant, and remove cross-coordinate covariance. Empirically, across five models, three retrieval domains, and 75 transformations, median component-count disagreement is 0.79 and median fixed-threshold decision disagreement is 0.26. A centering-only control isolates the reference-driven effect, with 1,141 of 1,200 component counts changing despite an unchanged observed spectrum, whereas independent parallel analysis seeds change none of the corresponding decisions. By contrast, orthogonally invariant comparator scores remain numerically stable with similar held-out discrimination. Together, these results show that parallel analysis-derived component counts and decisions can reflect hidden-coordinate choice rather than a well-defined property of the model.

High-frequency artificial phononic crystals offer a low-loss platform compatible with on-chip integration, yet realizing Chern phononic phases at GHz frequencies remains challenging. Here, we propose a magnon-induced phononic Chern insulator in a honeycomb phononic crystal hybridized with ferromagnetic islands at the hexagon centers. A circularly polarized Kittel mode couples to the surrounding phonons with a phase winding, which breaks time-reversal symmetry and opens a full Chern gap. In the large-detuning regime, this mechanism leads to an effective Haldane-type phononic model with magnon-induced complex hopping. By tuning the magnon-phonon interaction, the full hybrid system accesses Chern phases with tunable Chern numbers |C|=1 and |C|=2. The predicted gaps can exceed realistic phonon and magnon linewidths, enabling their observation in GHz acoustic devices. Our work establishes chiral magnon--phonon hybridization as a route to magnetically reconfigurable topological phononics.

In the present work an efficient border control management procedure is proposed. Compared to operational queue management systems, whose operations are based on mostly static data, the proposed work takes into account dynamic traffic conditions, thus enabling optimal performance, even in cases of uncertainty. To this end, we are proposing a multi-modal Artificial Intelligence (AI) framework, tailored to th needs of border control systems, which enables real-time queue prediction, management, and resource optimization. The novel proposed approach integrates heterogeneous data sources and presents them through a unified representation by employing Long Short-Term Memory (LSTM) networks for queue forecasting. Furthermore, it leverages Model Predictive Control (MPC) and scheduling optimization to derive actionable control policies, which in turn can be presented to border control officers. The proposed work has been evaluated using synthetic data simulating realistic traffic. The evaluation results demonstrate that the proposed method reduces queue prediction error by up to 35% and average waiting time by 30%. Accordingly, the average throughput increases by nearly 20%, compared to ARIMA and rule-based methods. The abovementioned results show the effectiveness and efficiency of combining AI architectures with optimization techniques for proactive and adaptive border traffic management.

Representation learning has attracted great atten- tion and managed to reach good performances as a pretraining method for downstream tasks or as a first step towards unsu- pervised speech modeling. Yet, little is known about how such methods deal with out-of-domain speech and how could they be adapted in a few shot to new domains. This is important especially for accented speech where one observes a long tail of accents that diverge from the standard ones. We introduce ABX- Accent, a benchmark based on the AESRC dataset that features 10 different accents of English. It includes a small (< 10 hours) unlabelled training set in each of the accents and adaptations of the Zero Resources Challenge ABX evaluation metrics to each of the accents. We illustrate this benchmark with a baseline model that uses adaptive domain normalization to fine tune a pretrained Contrastive Predictive Coding model on the accents. This method is first developed on LibriSpeech using a male/female split. When applied to the new benchmark, the proposed method yields a relative improvement of 23.6% on across-speaker ABX scores on average compared to non adapted models. The data and metrics will be open sourced upon paper acceptance

Powerful code agents can execute scripts, call tools, and manage files, yet many important applications remain accessible primarily through graphical user interfaces. We argue that screenshot-and-click is an inefficient interface for software-operating agents: screenshots are state-incomplete, and GUI actions are brittle, semantically weak, and poorly matched to long-horizon planning. We introduce ASIL (Agent-Software Interaction Layer), an agent-native interface that exposes software through structured JSON observations and code-executable semantic actions, realized through the deepest feasible access path for each application. We instantiate ASIL across 15 applications and a benchmark of 300 single-application and 80 multi-application tasks. ASIL reaches above 80 with closed models while executing fewer than five actions per task. Under a repaired runtime and a 50-step screenshot budget, the same tasks yield 6.6 and 26.6 strict success under screenshot-and-click control, rising to 15.0 and 53.3 on an easier OSWorld-comparable band. Against application-native interfaces on matched tasks, ASIL exceeds LibreOffice's UNO API by 28-38 strict points but only matches draw.io's MCP content contract. The structured modality also suits training: small-scale SFT raises Qwen3.5-2B from 58.0 to 72.1 and Qwen3.5-9B from 66.6 to 80.4, and resource-limited on-policy RL further raises them to 74.4 and 82.2.

Large language models can summarize financial information, but an operational stock-research system must first assemble heterogeneous evidence, expose unavailable data and model capabilities, and control how generated opinions affect a final report. We present DSA, an evidence-aware orchestration framework for multi-market stock research with large language model (LLM) agents. DSA organizes the workflow into evidence acquisition, structured context construction, model-routed analysis, optional role and Strategy Skill reasoning, and report generation with selected context and diagnostics. A default report profile and an optional agentic profile share evidence and model-routing services but use profile-specific output validation and risk safeguards. In the agentic profile, core role outputs are processed by role-specific parsers, whereas Strategy Skill opinions undergo an additional signal-eligibility partition before synthesis; disagreement is supplied explicitly to the decision agent, followed by a conservative risk override. The reference implementation includes six regional market paths, fifteen bundled Strategy Skills, hosted and local model routes, and multiple execution and delivery surfaces. At a frozen software snapshot, a selected manifest of 1,457 portable offline backend contract tests passed; 596 cases were retrospectively mapped to six contract families central to the reported LLM-agent architecture. This evidence establishes implementation conformance for the tested software contracts, not superior report quality, forecasting accuracy, or investment returns.

This paper investigates decentralized multitask learning over networks when the underlying task relationships are unknown. While existing graph-regularized multitask frameworks typically assume a known structure, practical settings often require learning inter-task dependencies directly from distributed data. We propose a decentralized two-phase strategy that first estimates a generalized graph Laplacian from noisy non-cooperative stochastic gradient iterates, and subsequently exploits the learned graph to enable cooperative multitask diffusion learning. This framework is motivated by a Gaussian Markov random field prior, which gives rise to a decentralized maximum likelihood estimator for the graph Laplacian. The analysis quantifies the Laplacian estimation error and its propagation to the steady-state performance of the multitask diffusion recursion, and introduces a topology sensitivity index to capture the effect of network heterogeneity. Simulation results corroborate the theoretical findings and demonstrate that cooperation enabled by the learned task graph significantly improves performance over non-cooperative learning, while approaching the true-graph baseline when the estimation stepsize is sufficiently small.

Organizing long-term memory for multimodal agents remains challenging because existing methods either suffer from expensive question-agnostic offline summaries or naive embedding similarity matching that introduces incomplete and redundant context. To address these issues, we propose GraphMemix, a combinatorial-optimization graph memory framework that models memory organization as query-aware evidence-forest construction. Specifically, our method consists of three key components:(1) candidate graph construction, which expands multi-view seed memories through schema and semantic relations to acquire query-aware original context; (2) evidence utility and activation costs, which decouples direct memory support from anchor-conditioned relation verification to suppress redundant or conflicting information; and (3) forest optimization, which jointly selects a forest-format memory context under a maximum evidence budget and its reliable relational structure. By organizing memory into a query-relevant subgraph, the method avoids substantial lifecycle cost and recovers low-similarity complementary evidence. Experimental results across four long-term multimodal memory benchmarks demonstrate significant improvements with different foundation models and establish a new Pareto frontier between accuracy and lifecycle cost.

Large language model (LLM) judges are increasingly used across various evaluation scenarios, making their judgment capabilities valuable intellectual property. However, black-box access exposes these capabilities to model extraction attacks. Existing extraction methods do not specifically target LLM judges and provide limited support for multiple evaluation protocols under restricted query budgets. In this study, we propose JUDGESTEALER, the first query-efficient model extraction framework for replicating judging capabilities across pointwise scoring, pairwise comparison, and listwise ranking protocols. JUDGESTEALER exploits the strong cross-protocol agreement to acquire pointwise scores and transform them into pairwise and listwise supervisions without additional victim queries. To capture informative judge patterns and improve query efficiency, JUDGESTEALER dynamically selects pointwise inputs based on semantic diversity, predictive uncertainty, and potential judge biases. It further applies score smoothing and multi-protocol review to preserve the ordinal structure of scores and mitigate catastrophic forgetting during surrogate adaptation. Extensive experiments on state-of-the-art LLM-as-a-judge and reward models show that JUDGESTEALER consistently outperforms existing extraction baselines, achieving up to 73.3%, 87.0%, and 71.6% accuracy for pointwise, pairwise, and listwise evaluation, respectively. JUDGESTEALER also remains effective across different sur- rogate model scales, adaptation strategies, and reasoning settings. Moreover, JUDGESTEALER demonstrates robustness against representative extraction defenses.

Landscapes are named, but whether names retain measurable environmental information beyond broad geographic structure is rarely tested. We analysed 3,757 Welsh settlements using a frozen, source-audited 24-element lexical framework, preregistered outcome-specific models and geographically structured validation. The central comparison contrasted 101 settlements carrying high-terrain elements (\textit{bryn} or \textit{mynydd}) with 139 carrying low-terrain elements (\textit{cwm} or \textit{pant}). High-terrain names occupied locations 24.4 m higher relative to their 2-km surroundings (95\% CI, 10.8--38.1 m; Holm-adjusted $p$ = 0.00137). The association remained positive across prespecified 1-, 2- and 5-km neighbourhood definitions and was reproduced using an independently produced elevation source (24.1 m; 95\% CI, 10.6--37.6 m). Adding terrain-name polarity to a non-lexical spatial and settlement baseline reduced geographically held-out mean squared error by 4.63\%, 6.22\% and 7.30\% under 10-, 25- and 50-km spatial blocking, respectively, although improvement varied among held-out regions. River-related names provided weaker, directionally consistent evidence, while the preregistered woodland model was non-estimable. Residual spatial structure, unresolved name language and the absence of independent external replication limit interpretation. Selected Welsh settlement-name categories therefore retain measurable information about present-day terrain within Wales, without establishing individual etymology, causal naming, historical environmental memory or transferability to other naming systems.

Kernels measure similarity or correlation in tasks such as regression and classification. The Gaussian kernel, other names of which include squared exponential and radial basis function kernel, is one of the most popular in Gaussian process regression. We argue that the Gaussian kernel is best avoided and should never be used as a default. The argument rests on two results demonstrating that the Gaussian kernel is extremely brittle. First, the Gaussian kernel gives rise to a conditional variance that is unrealistically small. If the variance is used to quantify predictive uncertainty, catastrophic overconfidence is almost inevitable. Second, a small variance goes hand in hand with numerical ill-conditioning, so that to use the Gaussian kernel in practice requires tricks such as nugget terms that effectively modify the underlying regression or classification model. These problems are caused by the unnatural smoothness of the Gaussian kernel, a fact we are far from the first to take notice of. The problem is not the Gaussian form itself but the analyticity of the kernel: Our argument is more broadly that analytic kernels are best avoided. For stationary kernels analyticity is essentially equivalent to an exponential decay of the spectral density.

Pre-training under limited data requires a different view of scaling than web-scale language modeling. With a fixed data budget but relatively abundant compute, increasing parameter count helps only up to an optimal scale; beyond that point, models overfit and generalization worsens. We study this behavior across 10M-100M word pre-training budgets, two corpora, and multiple downstream evaluations, and find that optimal size depends strongly on both the data budget and the downstream target. We argue that standard Transformers scale down poorly to this setting, because embeddings consume a large fraction of the parameter budget and per-token computation is tied to representational capacity. To address this coupling, we study recursive Transformers, reusing a shared block across depth to scale compute, together with factorized embeddings to reduce vocabulary-map parameters. We train three recursive models and find that they outperform standard Transformers at 10M and 100M words, while remaining competitive with BabyLM Challenge 2025 winners.

Place names provide persistent records of how landscapes have been described and organised, but their quantitative reuse requires explicit separation between mapped places, lexical annotations and environmental measurements. TEMPLAR Wales is a georeferenced environmental-toponymy dataset comprising 3,757 settlement records across Wales. The resource links a reproducible settlement frame to deterministic lexical screening and settlement-level environmental attributes through stable identifiers. It contains 1,350 lexical detections across 1,294 settlements, generated from a frozen registry of 24 Welsh place-name elements, while retaining exact- and prefix-token matches and their provenance separately. Environmental attributes describe river and coastal proximity, elevation and local terrain context at multiple spatial scales, land cover and neighbourhood woody cover, with parallel terrain measurements derived from independent elevation products. The dataset is distributed as four relational tables accompanied by a field-level data dictionary, source-provenance register and licensing metadata. Technical validation confirms relational integrity, deterministic lexical reconstruction, documented environmental coverage, strong agreement between independent terrain sources and reproducible reconstruction of the frozen release. TEMPLAR Wales provides a reusable foundation for research in toponymy, linguistic geography, historical and environmental landscape studies, GIS and spatial data analysis without treating computational lexical detections as verified etymologies or contemporary environmental measurements as historical landscape reconstructions.

This paper introduces ClusterAttention, a general training-free speedup of bidirectional attention layers. Existing sparse attention methods either rely on structure in the input, such as order in language or spatial proximity in images, or use slow clustering processes amortized over several forward passes. ClusterAttention instead uses a fast recursive clustering method that adapts to the geometry of the keys and queries in each attention head to produce useful clusters. This method allows setting the size of the clusters arbitrarily. We utilize this by setting all clusters to be a fixed size that is a power of two, allowing the block-sparse attention to run at the same latency per query-key interaction as dense attention on GPUs. We also derive an expression for the output error in sparse attention, that explains the counterintuitive experimental finding that tight clusters can lead to larger errors than random clusters. We then derive the error when excluded clusters are compensated through their centroids, and show that this error shrinks with tighter clusters. We integrate this compensation into the method. On large-scale tabular data ClusterAttention speeds up TabPFN-3 arXiv:2605.13986 by two to six times, while retaining at least 99% of the dense accuracy. To our knowledge, it is the first training-free method that can be successfully applied in the setting of unstructured input and a single forward pass. For video generation with Wan 2.1-14B T2V arXiv:2503.20314 , ClusterAttention achieves output closer to dense attention and a larger speedup (1.8x versus 1.4x) compared to SVOO arXiv:2603.18636 , a leading method developed specifically for this domain, both run without offline calibration.

12-lead electrocardiogram (ECG) is a standard, non-invasive examination widely used for diagnosing coronary artery disease, where clinical interpretation relies on comparing waveform patterns across multiple leads. However, most existing ECG analysis methods focus on single-lead signals or treat each lead independently, and typically process ECG signals as one-dimensional time-series data using CNNs or RNNs. While effective in modeling local waveform changes, such approaches have difficulty capturing inter-lead dependency and global waveform patterns essential for clinical diagnosis. To address this limitation, we propose a graph-based pseudo-multimodal contrastive learning framework called Graph-CMMC. ECG waveforms are transformed into Gramian Angular Difference Field (GADF) images to construct complementary representations of the same cardiac activity, enabling a pseudo-multimodal learning setting. Using all 12 leads, Graph-CMMC aligns waveform and GADF representations in a self-supervised manner, while a graph-based relational module is employed to model inter-lead dependency and enforce structural consistency across leads during contrastive learning. Experimental results on a multi-label coronary artery occlusion classification task demonstrate that the proposed framework achieves competitive performance compared to supervised learning methods. These results further suggest the effectiveness of using GADF as a complementary representation and incorporating explicit graph-based modeling of inter-lead dependency for learning robust 12-lead ECG representations.

This work concerns adversarial training against the small-norm adversarial examples that arise from the inherent input instability of a trained deep neural network. Examples in this class are small as measured in the relative $\ell^2$-norm, and therefore lie in the neighborhood of the input on which the model acts approximately linearly, the regime in which the perturbation remains imperceptible. We first show that such examples can be computed directly from the trained network parameters, without input gradient iterations, by means of a linearization called the low-rank Householder expansion (LRHE). The expansion describes the composed affine map rather than any individual layer, and the directions it identifies are read from the activation pattern already available in the forward pass. We then propose a simple adversarial training scheme built on this construction. No differentiation with respect to the input is performed at any point: training requires only additional forward evaluations, with weight parameters updated by the standard backward pass, and the inner maximization of the usual min-max formulation is eliminated entirely. That such a regularizer exists is our main finding: the methods that dispense with the inner search all obtain their local geometry by differentiating with respect to the input, and we show this is not necessary. The regularizer costs the equivalent of $2.8$ PGD steps per epoch, an $8.7\times$ reduction relative to 40-step adversarial training on MNIST and below the cost of 3-step training. The resulting models match three-step PGD adversarial training for relative $\ell^2$ budgets $\varepsilon \le 0.02$ and 40-step training for $\varepsilon \le 0.012$, falling away beyond, consistent with the locality of the expansion.

Molecular crystal structure prediction (CSP) is important in pharmaceuticals, agrochemicals, and organic electronics, where subtle differences in molecular conformation and packing can strongly affect material properties. We present Packora, a flow-based generative model for molecular CSP that jointly predicts atomic coordinates and the lattice from molecular graphs. Packora supports multi-component and organometallic crystals and can condition on any subset of molecular conformers, stereochemical labels, and space-group information within a single model. Inspired by the CCDC CSP blind test, we evaluate generation and ranking separately, using generation to isolate generator quality and ranking to measure end-to-end performance under a common relaxation and ranking pipeline. We also systematically study architecture, training, conditioning, inference, and scaling, identifying an effective design based on cacheable pairwise reasoning, training objective and numerical solver choices, conditioning dropout, and balanced scaling of pairwise and single representations. Packora outperforms the baselines on both structure generation and ranking benchmarks, achieving the best matched-budget coverage across all six generation benchmarks, as well as higher experimental-form recovery, lower experimental-form ranks, and faster convergence in ranking.

Graphs are invariant under node permutations, motivating the use of permutation-equivariant architectures in generative models. In flow matching, however, symmetry may also enter the source--target coupling: once graph pairs are compared up to node relabeling, the natural Wasserstein geometry is that of the graph quotient space. The Euclidean quotient metric of this space coincides with the Gromov--Monge distance, obtained by optimally relabeling the nodes. We develop this perspective theoretically, showing that quotient couplings can be lifted to aligned representatives without additional cost and that symmetrization yields equivariant flow-matching minimizers, including for categorical endpoint prediction. In practice, exact Gromov--Monge alignment is intractable, so we construct minibatch couplings using efficient Gromov--Wasserstein-type relaxations and lower bounds for the inner node alignment, optionally combined with an outer assignment between graphs. The resulting procedure changes only the training coupling and is compatible with standard permutation-equivariant architectures. Across continuous graph and categorical molecular generation, these structure-aware couplings substantially improve sample quality at small integration budgets, while our scaled-up molecular models remain competitive under conventional many-step sampling.

Scaling model-generated data is usually viewed as improving distillation: more examples should increase coverage, reduce noise, and produce stronger students. We show a second effect: larger datasets can make subtle teacher-specific signals easier to detect in the trained student, even when examples are off-task and never mention the trait. In a controlled setup inspired by subliminal learning, a teacher induced to express a target trait generates restricted off-task data, such as number-only completions. Students trained on different amounts of independent off-task data are evaluated in a separate domain, with matched no-trait controls isolating target-specific transfer. Our main finding is that larger independent datasets make the teacher's induced trait stand out more clearly in the student's later behavior. Other plausible traits may also strengthen with scale, but the target usually grows more. When the small-scale student already favors the target, scaling mainly amplifies that behavior; when it favors a related or salient alternative, more data can shift behavior toward the intended trait. Analyses of learned LoRA updates show a parallel trend. These effects appear across model families, trait types, multi-trait settings, and cross-model transfer. Our results suggest that scaling generated distillation data should be paired with trait-aware curation and evaluation, even when the data appears off-task or benign.

Security patterns are intended to support the design and development of secure software systems. However, although established catalogs of security patterns exist, their practical application remains limited. In particular, despite these catalogs, concrete patterns for common security controls such as user authentication (authentication for short) are lacking. This paper aims to make an initial contribution toward closing this gap, as exemplified by authentication. It presents a novel authentication pattern catalog, comprising 14 user authentication patterns. To support the catalog's practical application, it classifies patterns by the well-known concept of authentication factors and by the usual role each pattern fulfills in practice. By cataloging common authentication techniques through authentication patterns, we aim to make an important contribution to supporting software engineers and architects in designing and developing secure software systems.

Feed-forward 3D Gaussian Splatting reconstructs an explicit Gaussian representation from multiple input images in one network execution, making 3D reconstruction increasingly accessible for casual captures. However, such captures frequently contain transient objects that appear in only a subset of the views. Such content can be encoded into the per-view Gaussians associated with the inputs that observe it and remain in the combined representation despite being observed by no other input. As a result, it may produce blurred, duplicated, or floating artifacts in novel views. We introduce a training-free filtering procedure that exploits this per-view prediction structure. For each input, we exclude its associated Gaussians and render the same camera using the remaining representation, revealing content that is inconsistent with the other inputs. Feature similarity forms candidate regions, and rendering-based verification retains only candidates whose removal reduces reconstruction error in the other input views. The procedure operates on a single frozen prediction without retraining or scene-specific optimization. Across three reconstruction models and two distractor benchmarks, it consistently improves novel-view quality with varying numbers of input views. On clean scenes, evaluations across four models show that the original reconstructions are largely preserved.

Large Language Models (LLMs) are evolving from performing end-to-end mathematical reasoning to integrating agentic intelligence. However, most existing math benchmarks evaluate only final answers. This outcome-oriented evaluation provides limited diagnostic value for identifying process-level failures or rigorous logic, failing to guide the transformation of LLMs into robust agents. To bridge this gap, we present a process-level benchmark designed to evaluate the inherent agentic mathematical reasoning abilities of LLMs. Our framework aligns problem-solving agentic behaviors with a structured taxonomy of reusable mathematical atomic capabilities. We design a comprehensive suite of planning, action, and feedback tasks across both textual and multimodal contexts, supported by an automated pipeline that synthesizes high-quality trajectories and produces fine-grained annotations via controlled LLM rewriting. Experiments reveal that models with similar end-to-end accuracy can exhibit markedly different agentic capability profiles. This demonstrates that process-level evaluation is crucial for interpreting the true potential of LLMs and guiding the development of next-generation mathematical agents.

Answering questions over real-world documents requires processing long inputs that interleave text with tables. Optical context compression, which represents context as images, promises to reduce token cost, but its effect on table understanding remains unclear. We study pixel-level table compression for question answering over documents with multiple tables, evaluating five VLMs across two benchmarks and five visual-token budgets. Representing tables as images at native resolution matches text in both performance and efficiency, but downscaling them makes models compensate the loss in readability with longer, less effective reasoning traces that cancel the expected savings. Highly downscaled tables, however, preserve enough signal to identify whether they are relevant to a question. We exploit this asymmetry with a training-free, two-step method: the model first identifies the tables needed to answer a question from a pixel-compressed context, and then reasons over those at native resolution. On long documents, our method saves 41% of total tokens and gains 7 accuracy points over single-step QA with native resolution tables. It also uses 15% fewer tokens than the most efficient single-step compressed configuration, with no accuracy loss.

This paper proposes a novel data-driven algorithm to approximate the dominant eigenfunctions (aka.~modes) of the Koopman operator of nonlinear dynamical systems using neural networks. The relevance of learning the dominant Koopman modes is to approximate nonlinear dynamics by linear ones in a lifted space, thereby enabling simplified control and analysis. To fight the curse of dimensionality arising from using expressive templates (here neural networks) for the mode approximation, the proposed method leverages a power-iteration scheme that directly learns the dominant Koopman modes without explicitly constructing the projection of the Koopman operator on the template of functions. Our approach connects to other approaches in the literature that avoid the curse of dimensionality by learning small dictionaries of functions, but differs from them in that we do not require ``anti-collapse mechanisms'' to ensure that the learned dictionary is expressive enough to approximate the Koopman operator since our power-iteration scheme is designed to converge toward the dominant modes of the projected Koopman operator. The approach is fully data-driven, requiring only sampled state transitions. Theoretical guarantees are provided, showing convergence under increasing sample size and network width (in connection with the neural tangent kernel theorem). Numerical experiments demonstrate that the method achieves accurate and smooth approximations of dominant modes while avoiding the limitations of traditional techniques such as extended dynamic mode decomposition.

We present KinyaEmbed, the first dedicated sentence embedding model for Kinyarwanda, a morphologically rich Bantu language spoken by over 12 million people in Rwanda. Existing multilingual embedding models such as LaBSE, mE5-large, and OpenAI text-embedding-3-large perform poorly on Kinyarwanda due to severe under-representation in their pre-training corpora. KinyaEmbed is built on KinyaBERT-large and trained via a four-stage curriculum using MultipleNegativesRankingLoss (MNRL): Stage 1 leverages ~18,000 paraphrase pairs from the Official Gazette of Rwanda with three temperature scales; Stage 2 fine-tunes on 715 NLLB-translated MNLI triplets for entailment structure; Stage 3 aligns representations using English-Kinyarwanda OPUS-100 translation pairs; Stage 4 refines with 2,936 high-quality pairs filtered from KinyaCOMET at quality threshold 0.8. We evaluate on SemRel2024-rw and introduce Wiki-RW-STS, a new contamination-free Kinyarwanda STS benchmark of 300 pairs derived from Kinyarwanda Wikipedia. A seven-checkpoint ensemble (all5+23A*2, with the final stage double-weighted) achieves Spearman \r{ho}=0.7298 on SemRel2024-rw, surpassing mE5-large by 20.9% and OpenAI text-embedding-3-large by 41.0%. KinyaEmbed also achieves the best document clustering silhouette score (0.2146) across all evaluated models. All checkpoints, the KinyaCOMET filtered pairs, and the Wiki-RW-STS benchmark are publicly available.

Small language models (sLLMs) are nowadays hosted on devices with limited memory and computational budget. In an autoregressive setup, inference is memory-bandwidth bound: uniform quantization is often detrimental to such models, since their architecture has limited redundancies and only a few layers are not very sensitive to lower precision. We propose a composite metric that combines two orthogonal criteria: information retention (measured in terms of a normalized SQNR-based coefficient) and throughput gains (modeled using a roofline-based latency analysis). By profiling Gemma 3 1B, we find that Feed-Forward Network blocks and the embedding matrix are the most promising targets for acceleration. For each candidate, we estimate a normalized quality score based on simulated quantization and a normalized speed score based on roofline modeling with no actual execution needed. We combine the two scores in a composite priority coefficient, allowing us to tune the trade-off between speed and quality as needed. Our metric is general and can be used to prioritize individual blocks, their projection sublayers, or transformer layers as a whole. We evaluate our approach on several model architectures, showing that our estimates have at around 4% prediction error for the accelerated speedup. We find that our method generally allocates more resources to the most expressive layers compared to evolutionary search, specialized accelerators, or Shapley-value-based approaches that require expensive approximate inference. Our analytical approach makes sLLM quantization a predictable engineering task.

In many low-resource settings, even just eliciting speech for data collection is difficult. One promising approach has been to ask speakers to describe images. But how do we build models from such visually grounded speech data? Given a dataset of images with Hindi spoken captions, we consider how we can map a written English keyword to spoken realisations of that word in Hindi. Previous work trained end-to-end multimodal neural models. Instead, we explore a simpler alignment-based approach built on self-supervised speech representations. Written English tags are automatically obtained from images using off-the-shelf image captioning systems. Hindi utterances associated with the same keyword are then aligned (using self-supervised features), and alignment evidence is aggregated to identify recurring speech segments corresponding to the target word. Experiments evaluating keyword spotting and localization show that our alignment-based approach outperforms a previous attention-based neural model. We also show the benefit of incorporating negative examples during alignment. Our work demonstrates that cross-lingual word-to-speech mappings can be learned directly from visual grounding without transcriptions or explicit model training.

We present TabuLM, the first language model pre-trained on Kinyarwanda tabular data. Kinyarwanda is a morphologically rich Bantu language spoken by over 12 million people in Rwanda, yet lacks any dedicated tabular representation learning resource. TabuLM extends KinyaBERT-large, a two-tier morphological transformer, with additive row, column, and cell-type embeddings and a learned table-structure attention bias that sharpens same-row and same-column attention. Pre-training uses two new objectives: Masked Cell Recovery (MCR), which masks entire cells and forces reconstruction from row and column context, and Column Type Prediction (CTP), which predicts column semantic types from observed cell values. We pre-train on 172 Rwandan government tables (~35,000 cells) from NISR, RAB, REB, and MoH open-data portals, and introduce TabQA-kin, the first native Kinyarwanda table question-answering benchmark comprising 526 QA pairs across 31 tables and four question types. TabuLM achieves 62.0% exact match on TabQA-kin, outperforming KinyaBERT-large by 5.7 EM points and all multilingual baselines (mBERT 49.3%, XLM-R 50.0%) by 11.7-12.7 points. Analysis shows that structural table embeddings are most decisive for comparison and lookup questions, while morphological awareness provides complementary gains. Our code, data, and pre-trained checkpoint are publicly available.

We introduce AraMS-28k, the largest publicly released line-level dataset of genuine historical Arabic manuscripts, comprising 14 books, 3,043 pages, and 28,600 annotated text lines (27,971 main-text, 629 margin). Thirteen books are hand-copied manuscripts spanning three script traditions -- Naskh, Ruq'ah, and Maghrebi -- and one is a lithographed printed edition included to broaden format diversity. Each line is labelled as main-text or margin, and margin lines that have an unambiguous attachment point in the main text are further annotated with an insertion anchor, recovering the manuscript's true non-linear reading order at line-level granularity -- to our knowledge the first such annotation released for a historical Arabic manuscript corpus. Because reference transcriptions are fully vocalised while manuscript hands are typically undiacritised, we release both the raw diacritised transcription and a diacritic-normalised counterpart for every line. The dataset was constructed with RefLAM, a reference-grounded annotation pipeline that aligns multimodal-LLM OCR against independently sourced clean transcriptions and routes every line through human review, combining automatic verification with expert oversight. We describe the construction and quality-control process, present the annotation schema, report dataset statistics at both the corpus and per-book level, and provide baseline HTR results using Kraken and HATFormer, including a cross-script generalisation gradient from in-distribution pages to fully unseen books. AraMS-28k is released with page images, line-level annotations, and fixed train/val/test splits under CC BY-NC-SA 4.0 to support reproducible research on Arabic manuscript recognition, layout analysis, and reading-order recovery.

Automating prostate radiotherapy treatment planning is dosimetrically complex, particularly for extreme hypofractionated regimens. In this study, we introduce Dose-PlanNet, a physics-guided 3D deep learning architecture designed to predict dose distributions. This model's performance was evaluated on a cohort of patients treated in a prospective trial where two different dose fractionation regimens were employed. Dose-PlanNet achieved comparable target coverage ($D_{95}$), though statistical analysis revealed a marginal reduction in target homogeneity ($p<0.001$) offset. However the model achieved statistically significant improvements in high-dose organ-at-risk sparing ($p<0.001$). When evaluated against strict Prospective Randomized protocol volumetric constraints, automated plans met prespecified clinical acceptance criteria in $11$ out of $14$ Moderate Hypofraction Arm plans and $9$ out of $12$ Stereotactic Body Radiation Therapy Arm plans. This pipeline demonstrates that physics-informed deep learning can accelerate radiotherapy workflows while safely maintaining the stringent dosimetric quality required for high-precision clinical deployment.

Automated candidate-job matching systems are increasingly classified as high-risk AI under emerging regulation, yet auditing them for demographic bias is expensive: classical correspondence-audit studies require hand-crafted resumes and manual submission, which does not scale to fast pipeline retraining cycles. This paper presents a general, reusable methodology that (1) uses task-specialized LLM agents to synthesize identity-neutral base resumes and inject controlled demographic treatments across five protected-characteristic axes (sex/gender, age, residence, language, disability), producing a K x (1+N) correspondence-audit matrix; (2) qualitatively flags inferred protected characteristics per an EU AI Act-aligned prompt; (3) ranks candidates against a job description via a fine-tuned sentence-embedding model and cosine similarity; and (4) computes a nine-metric fairness suite spanning counterfactual (score delta, mean absolute rank change, flip rate), group-fairness (top-K retention, four-fifths/impact ratio), and merit-aware (Recall@K, nDCG@K, equal opportunity, equalized odds) families, each with bootstrap confidence intervals, significance tests, and Benjamini-Hochberg correction, culminating in an automated PASS/INVESTIGATE/FAIL report with a composite risk score. On an example corpus of 5 job orders, 100 base candidates, and 10 demographic treatments (90 metric x variant evaluations): score shifts, top-K retention, and merit-aware rate gaps stay within tolerance for every treatment, but a rank-stability metric (MARC) and nDCG@K each surface borderline findings - including one on the neutral baseline itself - that a score- or retention-only view would miss. The results argue for multi-metric, multi-family auditing over any single aggregate score, and for LLM-agent-generated audits as a practical, low-cost complement to human-curated audits for any candidate-job matching pipeline.

As electricity market participants increasingly adopt learning-based agents for their bidding strategies, electricity markets are becoming algorithmic. Evidence from algorithmic markets in other domains shows that tacit collusion can arise purely through independent learning. Moreover, electricity markets are typically oligopolistic and feature repeated interaction among a small number of participants, making them structurally susceptible to non-competitive behavior. In the face of these observations, this paper investigates the hypothesis that tacit collusion may emerge in electricity markets where participants' actions are controlled by autonomous learning-based algorithms. We model strategic bidding as a repeated game with imperfect public monitoring, and model the participants' emergent behavior using multi-agent reinforcement learning. We propose a multi-dimensional set of criteria (going beyond profit comparisons against Nash equilibria) to assess whether the resulting behavior constitutes tacit collusion. Our experimental results showcase that such a danger is realistic for electricity markets: there are cases where agents do learn to sustain supra-competitive outcomes that are supportive of tacit collusion indicators, even though the agents were never instructed to collude.

Agentic recommender systems ground each decision of a large language model (LLM) in a persistent memory of the user, and in existing agents that memory is text: a narrative written and maintained by further LLM calls. Text limits this memory in two ways. It is updated one rewrite at a time, so exploiting the full interaction history is prohibitively expensive; and collaborative evidence, graded similarity over an entire catalog, does not survive translation into sentences. We propose CoVeMem (Collaborative Vector Memory), which vectorizes the collaborative core of the agent's memory. Frozen LightGCN user and item states form the memory bank; at each decision, the candidate set itself retrieves the most relevant historical states, which enter the LLM's context as soft tokens alongside a light textual profile. Contrastive alignment to item-semantic anchors, followed by listwise co-training with masked candidates, teaches the model to read these states and to rank through them; a pointwise yes/no readout scores each candidate. Across four instruction-grounded recommendation benchmarks, CoVeMem matches or exceeds the strongest collaborative text-memory agent on 19 of 20 metric cells while requiring zero additional LLM calls for memory maintenance beyond the shared static profile, against per-interaction calls for text memory. The memory now takes gradients: the full interaction history, out of reach for text, becomes available as training data for what the agent remembers and for how it reads what it remembers.

Learning-augmented algorithms improve online decisions using predictions, but unreliable advice may harm efficiency and fairness. We study an online allocation problem with finite candidate sets, irreversible decisions, and exposure constraints. We propose a robust and fair rule combining advice with a conservative fallback and fairness correction. Under bounded-error assumptions, we prove consistency and robustness with loss proportional to prediction error. Experiments show stability under adversarial advice and significant reductions in exposure disparity.

LLMs are thought to track "belief states," i.e., running probability distributions over the latent variables that govern language (Shai et al., 2024; Sarfati et al., 2026), but so far this has only been comprehensively demonstrated on toy synthetic data and in a few isolated case studies. It has also never been empirically connected to the geometry of LLM features (the concepts interpretability finds in model activations). In this work, we plant a controllable latent variable inside natural-looking text. An LLM teacher writes ordinary text while we "subliminally" steer it along one of K = 8 unrelated sparse autoencoder directions at each token, with the active directions following a ring-shaped Markov chain. A small transformer model trained on this corpus does indeed track the Bayesian posterior belief about our planted latent variable. Moreover, it also arranges the 8 states themselves on a ring, in the exact order of the Markov chain, which is supporting evidence that a concept's geometry can be formed by the statistical dynamics of the latent variable behind it.

Background. Large language models (LLMs) are increasingly used for screening in evidence synthesis, where false negatives can remove relevant studies before full-text assessment. We compared human and LLM title-and-abstract screening workflows in a preregistered study embedded in a conceptually complex scoping review. Methods. After a conservative title-only screen, 1,131 records were screened by one review lead, four trained assistants screening non-overlapping subsets, and seven complete LLM runs using different models and processing configurations, including a nominally identical repeat run. We compared retained workload, operational recall against 316 verified eligible records, agreement, run-to-run consistency, and procedural burden. Because eligibility was verified only for records advanced and assessed in the parent review, recall estimates were operational. Results. No workflow recovered all verified eligible records. The human workflows and two GPT-5.4 file-batch runs retained 42.2-45.0% of records while achieving 82.3-82.9% recall. Gemini 3.1 file batches achieved the highest recall (83.9%) but retained 56.7% of records. All-at-once configurations recovered fewer eligible records than corresponding file-batch configurations. Two nominally identical GPT-5.4 file-batch runs agreed on 91.7% of records but differed on 94 records, including 29 verified eligible records retained by only one run. Discussion. LLM screening performance depended on the implemented workflow, not model identity alone. Processing configuration, workload, record-level variation, and human-LLM decision integration are therefore substantive properties of deployed systems. For high-recall tasks, LLMs are better suited to validated, auditable, human-supervised workflows than autonomous exclusion.

Industrial control systems (ICSs) rely on programmable logic controllers (PLCs) to connect networked computation with physical control. Tool-using large language model (LLM) agents represent an emerging attack threat: can an autonomous agent convert a network-reachable PLC into sustained adverse physical impact? However, existing evaluations focus on digital tasks or individual stages of PLC testing. In ICSs, evaluations that stop at software exploitation, an accepted write, or tool access may therefore mischaracterize physical risk. We present PLCBENCH, to our knowledge, the first real-PLC hardware-in-the-loop (HIL) framework for characterizing this cyber-to-physical capability and its boundaries. It combines vendor-native interaction, commercial PLC execution, closed-loop reduced-order process simulation, and independent outcome verification. A deterministic evaluator applies fixed rules to runner, communication, PLC-object, and process records to assign six hidden diagnostic flags, distinguishing usable PLC interaction, process-linked manipulation, and sustained physical impact. We instantiate PLCBENCH on four commercial PLCs crossed with four closed-loop workloads. Across five LLM families and 240 real-PLC episodes, 75 episodes (31.3%) sustain their respective physical objectives. Stagewise results show that 98 episodes stop before a valid native read, whereas 62 reach a process-linked write but do not sustain the final objective. Notably, richer process observation is associated with an increase in conditional objective attainment after a process-linked write from 44.2% to 64.0%. These measurements localize failure in configured PLC-process deployments and identify intervention points for future defense evaluation. To support reproducibility, we release the safely disclosable PLCBENCH code and a software-only reproduction pipeline through the accompanying artifact.

Fusing multiple modalities is expected to improve model performance. However, on the MultiHuSE dataset, early, late, and symmetric attention fusion often fail to outperform the best unimodal baseline (text). Pathway isolation of a symmetric attention fusion model reveals that the text-pathway accuracy drops from 74.9% to 56.4% after fusion in one such setting, indicating that the dominant modality can be degraded during integration. We term this strong-modality collapse and argue that it helps explain why some multimodal models fail to surpass unimodal baselines. We propose Inverted Asymmetric Fusion (IAF), which avoids forcing mutual attention across modalities. The dominant modality is preserved by passing through fusion unchanged, while weaker modalities attend to it as a contextual anchor. Before fusion, weaker modalities are strengthened using Modality-Aware Knowledge Distillation. We evaluate IAF on three benchmarks with different modality hierarchies: text-dominant datasets (MultiHuSE, UR-FUNNY) and an audio-visual-dominant dataset (MUStARD). Pathway isolation shows that IAF preserves the dominant modality's internal accuracy at its unimodal ceiling across all tested configurations, whereas symmetric fusion degrades it by up to 18.5% on MultiHuSE. IAF improves over the strongest unimodal baseline by up to 8.25%.

Prior analyses by Derezinski and Warmuth established all-size sampling identities, selected-OLS unbiasedness, and inverse moments for ordinary volume sampling, while their exact arbitrary-fixed-response loss and prediction-covariance formulas are at the rank-size endpoint s=d. We establish a Loewner envelope for centered coefficient covariance for every full-rank fixed pool, response, and legal budget d <= s <= m under ordinary indexed fixed-size volume sampling followed by selected unweighted least squares; its coefficient is globally sharp over the full-rank class. Global sharpness does not determine attainability on the pool in hand. Under positive loss, strict-interior budgets, and no coloops, a feature-only margin nu_A gives the exact fixed-design spectral phase: nu_A > 0 if and only if the normalized spectral envelope is strict for every compatible residual, whereas nu_A = 0 if and only if some compatible residual is spectrally tight; the same zero-margin residual is tight at every strict-interior budget. A residual-augmented change of measure supplies the response-aware mechanism and a one-sided quantitative slack bound, while support saturation proves the attainment direction. Critical equal-leverage geometry interprets the boundary, and sound lower certificates yield conservative same-primitive cardinality decisions. Frozen-feature examples show that the certificate is nonvacuous and measure the fixed-pool cost of its authorized reduction. The claims concern conditional centered, full-Gram-whitened coefficient covariance, not population generalization.

Weak signals are early, low-visibility indicators that precede significant changes before those changes become established. Existing detection methods, based on keyword frequency, topic modeling, or untyped graph topology, fail to capture the semantic and relational structure through which such signals manifest. In this paper, we propose C-Unseen, a self-interpretable framework for weak signal detection in Dynamic Temporal Knowledge Graphs (DTKGs). We define a weak signal as a rare, semantically coherent subgraph that proliferates across consecutive TKG snapshots. The framework operates through two modules: a Rare Subgraphs Extractor, in which an LLM identifies subgraphs whose content is in tension with the dominant snapshot narrative via chain-of-thought reasoning, and a Weak Signal Alerter, in which the persistence of these rare subgraphs is tracked across time steps to isolate true weak signals. Experimental results demonstrate that C-Unseen outperforms keyword-, topic-, and graph-based baselines.

Large language model agents reason, call tools, and act autonomously over many steps, but their agentic skills-correctly sequencing tools, planning under dependencies, judging untrusted inputs, and grounding generated arguments-are hard to measure with accuracy-only leaderboards. We present BekchiAI, which addresses both sides: a benchmark for measuring agentic skill and a platform for observing and controlling live agents. The BekchiAI-Benchmark, a suite of 13 tool-using ReAct agents across 7 task categories (arithmetic, structured/SQL, security detection, URL grounding, planning, orchestration, and tool-policy), totalling 2,057 deterministic, committed test tasks. Every task is verifier-checkable gold answers are computed by running canonical SQL against a real database, computing the exact schedule of a directed acyclic graph (DAG), or evaluating closed-form lambdas including adversarial security samples paired with deliberately imperfect signature scanners so a score reflects the model's own judgment, not the copying of an oracle. We define a small set of behavioral metrics beyond accuracy-tool-call adherence, URL hallucination and source-match, and per-model token cost and report a four-model comparison (Qwen3.7-Max, gemma-4-31B-it, gemma4:26b, gpt-oss-120b) whose story is in the per-family spread, not the aggregate. The benchmark runs are executed using the provided evaluation scripts. BekchiAI-Platform is a complementary web-based observability and control layer for deployed agents, providing full token and latency telemetry as well as remote run termination. The benchmark, evaluation tools, and platform are publicly released.

Connected and automated vehicle (CAV) platooning offers a promising approach to improving road safety and traffic capacity. However, platoon control in real-world traffic is challenging due to uncertainty and heterogeneous driving behaviors. Reinforcement learning (RL) has strong potential for addressing such control problems, but its practical deployment raises challenges related to safety and learning efficiency. This paper proposes a generic modeling and simulation framework for investigating CAV platoon joining maneuvers and comparing deep reinforcement learning (DRL)-based control algorithms. The problem is particularly challenging in mixed-traffic environments, where CAVs coexist with human-driven vehicles exhibiting heterogeneous longitudinal and lateral behaviors. The objective is to achieve safe and efficient joining maneuvers by either incorporating penalties for risky behaviors into the learning process or using an external safety controller to constrain the learned policy. An agent-based modeling framework coupled with the Simulation of Urban MObility (SUMO) simulator is used to evaluate Deep Q-Network (DQN), Double Deep Q-Network (DDQN), and Proximal Policy Optimization (PPO). Results show that PPO outperforms DQN and DDQN, achieving a joining success rate of approximately 98 % and a collision rate below 1 %, largely due to risk-related penalties incorporated into the reward function. However, this improved performance requires more decision steps to complete the maneuver, revealing a trade-off between safety, joining effectiveness, and decision efficiency. An external safety controller effectively prevents collisions, although its interventions may reduce joining efficiency. The results highlight the importance of jointly considering safety and efficiency when designing RL-based controllers for CAV platoon joining in mixed traffic.

Although Multimodal Large Language Models (MLLMs) have demonstrated impressive performance in Medical Visual Question Answering (Med-VQA), their reliance on global image features often lacks precise pixel-level grounding, thereby limiting clinical trustworthiness. To bridge the semantic gap between high-level clinical reasoning and spatial localization, we propose \textsc{\textsc{MedREAL}} (\textbf{Med}ical \textbf{RE}asoning-driven \textbf{A}nswering and \textbf{L}ocalization), a unified framework that seamlessly aligns linguistic reasoning with spatial grounding. Specifically, \textsc{MedREAL} introduces \textbf{S}eg \textbf{A}nchored \textbf{R}easoning \textbf{P}ooling (SARP) to distill task-relevant semantic evidence directly from \texttt{[SEG]} tokens within the MLLM's hidden states. Furthermore, a \textbf{R}easoning-to-\textbf{V}isual (R2V) fusion mechanism is proposed to effectively inject these reasoning-aware features into a segmentation pipeline for accurate mask decoding. To facilitate this paradigm, we construct MedRAVS-13K, a comprehensive dataset comprising 13,824 expertly validated samples across four diverse imaging modalities. Extensive experiments demonstrate that \textsc{MedREAL} significantly outperforms state-of-the-arts, achieving 68.49\% gIoU and 70.47\% cIoU on benchmark evaluations. By generating evidence masks that are strictly consistent with textual diagnoses, \textsc{MedREAL} provides a robust, interpretable framework for reasoning-driven medical image analysis.

User behavior simulation with large language models~(LLMs) is increasingly used to support multi-agent ecosystem simulation. Existing simulators typically rely on static user profiles inferred from historical observations, which become inadequate in socially intensive environments such as live streaming where interaction dynamics continuously reshape user behavior. We propose \textbf{LiveSim}, an LLM-based framework for live-stream ecosystem simulation. It represents users as editable behavioral hypotheses and progressively refines them through trajectory-grounded interactions, where discrepancies between simulated and observed trajectories reveal missing environmental shaping effects. These signals are further extracted as transferable environment-behavior patterns and accumulated in a collective behavioral memory to improve user-level behavioral fidelity and support ecosystem-level simulation. Experiments on real-world live-stream risk-control data validate the effectiveness of LiveSim in improving user-level behavioral fidelity and enabling ecosystem-level analysis of risk evolution and platform intervention effects.

Medical Visual Question Answering (Med-VQA) holds significant promise for clinical decision support, yet faces challenges due to limited annotated data and the high computational demands of existing large vision-language models. We propose MedFG-VQA, a lightweight framework that leverages a memory bank to augment DCT-based low-frequency features and employs graph-enhanced cross-attention for effective visual-textual alignment. Specifically, our approach features two key components: Frequency-Memory Fusion (FMF), which enhances low-frequency features by retrieving from a learnable memory bank built on DCT decomposition, and Graph-Aware Cross-Attention (GACA), which aligns visual-textual features via cross-attention and refines them through graph-convolutional aggregation. To address data scarcity, we construct SynMed-VQA, a large-scale synthetic dataset comprising over 2 million question-answer pairs across 9 imaging modalities and 10 major organs, generated with GPT-4o. Extensive experiments on SynMed-VQA and three other standard biomedical VQA benchmarks demonstrate that MedFG-VQA achieves competitive or superior performance compared to much larger models while maintaining significantly lower computational costs, highlighting its efficiency and potential for clinical deployment.

Interactive language-model agents use confidence signals to decide whether to answer immediately, retrieve additional evidence (from memory or external knowledge), or defer. Yet confidence is usually evaluated in isolation, without measuring the trajectory-level consequences of the actions it triggers. We propose matched trajectory replay, a controlled protocol for comparing confidence-to-action mappings. The protocol holds candidate answer states, evidence points, budgets, and action costs fixed. We use it to compare raw verbalized confidence with post-hoc isotonic calibration in a multi-hop question-answering system using Mistral, GPT, and Qwen models on HotpotQA and MuSiQue datasets. At the same numerical commitment threshold, calibration changes which questions agents ultimately commit to answering. Across all six model-dataset pairs, it increases accuracy among committed answers by up to 41 percentage points. However, it can reduce coverage and increase retrieval use. Overall accuracy improves by up to 15 percentage points on HotpotQA but falls by up to 17 percentage points on MuSiQue. These effects reflect a shift to a more selective, lower-risk operating point, not improved answers or confidence ranking. A calibration map fitted before retrieval improves held-out calibration through retrieval depths one and two, but is worse than raw confidence at depth three for all three models. Additional evidence helps on average, but this aggregate effect does not establish whether confidence identifies which individual episodes will benefit from another retrieval. Taken together, these results show that calibration can make commitment risk interpretable, but it does not estimate the expected benefit of another retrieval. Retrieval therefore requires a separate value-of-information or utility estimate. Evaluations should report held-out calibration, risk-coverage, and retrieval cost.

Large Language Models (LLMs) have demonstrated remarkable proficiency in natural language understanding, yet they struggle with strict multi-step reasoning, frequently suffering from hallucinations and inconsistency. Existing solutions like Chain-of-Thought (CoT) lack rigorous verification mechanisms, while standard Retrieval-Augmented Generation (RAG) often misses the complex, structural dependencies inherent in logical tasks. To bridge this gap, we propose a Neuro-Symbolic architecture that integrates a Logical Knowledge Graph (LKG) with dynamic solver routing. Specifically, we introduce an ontology-based LKG that treats logical rules and constraints as first-class topological nodes, enabling explicit modeling of dependencies extracted from text. We further design a Logic Router to dynamically dispatch tasks to the optimal symbolic engine, which is supported by a topology-aware hybrid retrieval mechanism. Experimental results on logical reasoning benchmarks demonstrate that our framework significantly outperforms state-of-the-art prompting and RAG baselines, delivering higher accuracy and verifiable reasoning paths.

Large language models (LLMs) are increasingly applied to specialized domains, where effective use of domain expertise often requires reasoning over complex rules in concrete scenarios. However, existing benchmarks only partially evaluate this capability, as they either focus on output-level instruction constraints or overlook the distinct roles that rules play in scenario reasoning. To address these gaps, this paper introduces RuleWeaver, a benchmark construction framework for evaluating rule-centered scenario reasoning. RuleWeaver starts from corpus-derived IF-THEN Meta Rules, progressively augments them into complex rules, and composes these rules into rule-centered scenario QA instances. Beyond final-answer correctness, RuleWeaver further supports process-level evaluation through rubric-based answer quality, rule recall, and rule precision. Experiments on 11 representative LLMs show that current models still struggle with complex rule-centered scenario reasoning, with even the best-performing model achieving only around 50% of the maximum rubric score. We make our code and dataset available here: https://github.com/SharkSpicy-NLP/RuleWeaver.

Time series forecasting models operate on raw numerical sequences, lacking the semantic knowledge that domain experts implicitly leverage, such as the physical meaning of each variable, its statistical behavior, and its temporal dynamics. Recent efforts to bridge this gap fall into two camps. Some rely on large language models at inference time, which is computationally expensive. Others apply uniform textual prompts at the dataset level, ignoring the heterogeneous semantics across individual variates. We propose SAGE (Seeing and Augmenting with Grounded Encoding), an end-to-end CLIP-based framework that jointly models temporal, cross-variable, textual, and visual information. The CLIP text encoder processes frequency-enhanced patches and variable tokens, while gated residual paths inject variable-specific descriptions and statistical descriptors. In parallel, the frozen CLIP vision encoder aligns rendered series with temporal representations through a training-only contrastive objective. This dual use of CLIP adds complementary semantic and visual supervision without placing an LLM in the forecasting loop. Across eight long-term benchmarks and M4, SAGE achieves state-of-the-art accuracy. Ablations confirm complementary gains from multimodal alignment and variable-level knowledge.

The National Oceanic and Atmospheric Administration (NOAA) employs independent prediction systems for distinct forecast products. While some separation is practical, we argue that combining short- and medium-range weather into a single prediction system would provide the public with a useful distillation of global weather and its impacts. To this end, we present Nested-EAGLE (Experimental Artificial intelligence Global and Limited-area Ensemble): a 0.25° global weather model with a 6 km refinement over the Contiguous United States (CONUS). The model achieves significantly lower mean-squared error in near-surface and low-level quantities over CONUS compared to NOAA's Global Forecast System and High-Resolution Rapid Refresh (HRRR), while remaining competitive throughout the rest of the global atmosphere. We show that the skill gains for near-surface fields stem from incorporating high-resolution regional analysis data into training through the nesting process. Forecasts of precipitation amounts are less skillful than those from HRRR, owing to deterministic training. However, we show that Nested-EAGLE provides the most accurate forecasts of storm locations at longer leads, despite blurred extrema. Our results motivate future work to extend the skill gains beyond CONUS and improve precipitation representation.

A novel Hyperspectral diffusion Equivariant Imaging (HyDiff-EI) framework for solving the hyperspectral image (HSI) inpainting problem has been presented here. Unlike conventional diffusion-based methods that rely on large-scale pretraining, HyDiff-EI is a test-time optimization framework that learns directly from a single corrupted HSI acquisition. This makes it flexible for different sensor configurations and particularly well-suited for practical remote sensing scenarios where large annotated hyperspectral datasets are limited. To address the ill-posed nature of unsupervised inpainting, we embed equivariant consistency constraints within the diffusion process. By leveraging the inherent geometric symmetries and intrinsic characteristics of HSIs, HyDiff-EI bridges the gap between generative diffusion modeling and self-consistent physical priors. We empirically show that coupling diffusion modeling with equivariant priors substantially enhances noise robustness and generalizability. Extensive experiments on real-world datasets including Chikusei, Botswana, and EMIT demonstrate that HyDiff-EI offers remarkable inpainting quality over existing self-supervised and diffusion-based algorithms in both noiseless and noisy cases.

Travel planning agents assist users in generating personalized travel plans by modeling their individual preferences. Existing agents either rely on explicit user instructions or engage in multi-turn clarification to elicit user preferences. However, both approaches overlook the rich behavioral signals latent in users' past behaviors, which implicitly encode their preferences. This over-reliance on active user input increases interaction burden and limits plan personalization. To bridge this gap, we introduce a new task, Behavior-Aware Travel Planning, which infers user preferences directly from past behaviors and generates personalized travel plans. To facilitate research on this task, we introduce Behavior2Trip, a benchmark constructed from one of the largest Chinese online travel platforms, comprising 11,400 instances. Each instance represents an average of 39.8 past user behaviors spanning 14 attributes across 5 preference dimensions. We further propose B2T-Agent, a reinforcement learning-based agent that leverages user behavior trajectories, interacts with external tools for preference-aligned retrieval, and maintains an internal memory module. Experiments on Behavior2Trip show that GPT-4.1 achieves a full-constraint pass rate of only 0.5\% on the hardest tasks, while B2T-Agent built upon Qwen3-8B outperforms all baselines, highlighting the substantial challenge of this task. Moreover, Qwen3-8B trained with B2T-Agent also outperforms GPT-4.1 on the TravelPlanner benchmark, demonstrating strong generalization. Code and data are available at https://github.com/BUAA-IRIP-LLM/Behavior2Trip

Recommendation impressions are a finite resource, hence delivering a recommendation to a user who would discover the content organically yields no incremental value and displaces other recommendations that could. We address this by extending an existing production recommendation model to a causal architecture using holdback data that is already collected as part of routine experimentation infrastructure, requiring no new data collection. A central challenge is that attribution windows differ between treated and holdback observations: treated users are attributed a stream within a short direct-response window, while holdback users are attributed organic streams over a multi-day window. This mismatch makes naive treatment-effect subtraction invalid. We resolve this with a dual-threshold targeting policy that delivers a recommendation only when the probability of a treated stream is high and the probability of organic stream is low. In a production-scale A/B test on millions of Spotify users, this policy reduces recommendation impressions by 7% with no statistically significant reduction in overall recommended content consumption. We further show that joint training with holdback data improves calibration of the treated head relative to the production baseline, and argue this can be taken as evidence that causal models learn more generalisable representations than models trained on observational data alone.

The rapid improvement of LLMs has made distinguishing AI-generated text from human writing a pressing problem. This challenge is further amplified by paraphrasing tools designed to make machine-generated text appear more "human". We study how access to human writing samples can be used to strategically paraphrase machine-generated responses toward the human distribution. Under a multi-sample setting with human and machine responses to the same prompts, we show that repeated paraphrasing moves the machine distribution toward the empirical human distribution under simple mixing and stability conditions. Our results derive an explicit convergence rate, extend the analysis to a finite-sample setting, and characterize how the required number of human samples and paraphrasing rounds scale with the desired error.

Recent work shows that pre-trained, instruction-tuned vision-language models (VLMs) perform well at mapping from instructions and observations to high-level plans, but struggle to realize such plans as reliable low-latency action sequences in unfamiliar environments. At the same time, world-model controllers excel at fast observation-to-action control, but lack open-ended task guidance. In this work, we combine these strengths into a single system, Instruct-to-Act, where we train a world-model controller to act autonomously at high frequency when conditioned on sparse, higher-latency, and high-level text instructions generated by a VLM planner. To train controllers to be language-instructable, we relabel segments of controller policy rollouts with synthetic instructions and jointly optimize a behavior-cloning objective along with existing reward-maximizing and world-modeling objectives. We evaluate our proposed approach across seven embodied environments, including three multi-agent environments where VLM planners coordinate through language while trained controllers serve as their actuators. Under matched observation and action spaces, our decoupled approach consistently outperforms controller-only and direct VLM action-generation variants, preserves fast control, and lets us swap in different pretrained VLM planners without fine-tuning, while remaining competitive with strong vision-language-action and multi-agent RL baselines on six of seven tasks.

Control system design is critical for modern industry, such as chemical process temperature regulation and aero-engine control. However,traditional control design workflows rely heavily on expert knowledge and extensive manual parameter tuning, resulting in limited efficiency and scalability. To this end, this paper proposes AI Control Scientist (AICS), the first large language model (LLM)-driven agent capable of automatically generating optimized controller from language design requirements. Specifically, a Task Modeling Agent interprets user requirements to engineering constraints; a Controller Design Agent generate candidate controller structures and executable code; and a Parameter Tuning Agent refine controller parameters under closed-loop performance criteria. Experiments demonstrate that the proposed agentic system can automatically generate multiple representative control systems, outperforms existing automated baselines in both design success rate and optimization efficiency. This work has the potential to transform control system design from human-driven to agent-driven, paving the way for model predictive control and other advanced control systems design.

Preference learning optimizes models using response pairs, yet the informativeness of these pairs is fundamentally shaped by the instructions from which they are generated. We identify instruction quality as a hidden bottleneck in preference learning: low-quality or ambiguous instructions restrict the response-quality distribution, limiting strong chosen responses and weakening preference signals. Through Best- and Worst-of-N analyses, we show that instruction quality constrains both the ceiling and floor of sampled response quality. Motivated by this observation, we introduce an instruction-refinement pipeline that selects weak instructions using reward signals and revises them with rubric-guided LLM feedback, improving preference data without discarding examples. Across offline and online preference learning settings, experiments on multiple models and benchmarks show broad alignment improvements over original data and alternative data-improvement strategies. Further analyses indicate that instruction refinement raises achievable response quality and complements response-centric preference data curation. Overall, instruction quality emerges as a key factor governing how informative preference signals are formed for LLM alignment. Code is available at: https://github.com/01choco/instruction-refinement/

Existing constant-step analysis of stochastic \Scaf{} identifies a leading $O(γ/N)$ stationary mean bias and shows that higher-order bias can persist as the client count increases, but does not identify the first client-independent contribution at coefficient level. For full-participation stochastic \Scaf{} with one-dimensional homogeneous clients, fixed local-step count $H$, and bounded additive gradient noise, we prove, uniformly over $N\ge2$, $$ \begin{aligned} \mathbb{E}_{π_{γ,N,H}}[x]-x^\star ={}& -\frac{f'''(x^\star)σ^2}{4f''(x^\star)^2}\fracγ{N}\\ &- \frac{f'''(x^\star)σ^2}{12f''(x^\star)} \frac{(H-1)(5H-1)}{H}γ^2 +O_H\!\left(\frac{γ^2}{N}+γ^3\right). \end{aligned} $$ Hence client averaging suppresses the leading $O(γ/N)$ bias but does not remove the client-independent $O(γ^2)$ component when its coefficient is nonzero. The mechanism is indirect: although the direct control contribution cancels pathwise in the linear global average, the controls still alter within-round local trajectories and their second moments. Fresh gradient noise and persistent control fluctuations therefore generate local second-moment corrections that nonquadratic curvature converts into stationary mean bias. The coefficient vanishes for quadratic objectives. Numerical experiments are consistent with the predicted coefficient, its persistence as client count increases, and the stated joint remainder. The result is restricted to the one-dimensional homogeneous fixed-$H$ setting.

Machine learning offers a possible route to data-driven real-space renormalization when the relevant observables are nonlocal and difficult to prescribe explicitly. We explore this idea for two-dimensional site percolation developping a supervised, scale-shared neural architecture. The model recursively applies the same learned coarse-graining rule across scales, producing a latent field from which the crossing probability is predicted, while a corresponding fine-graining decoder reconstructs the largest-cluster mask. Trained only on small lattices, the model extrapolates to substantially larger systems, recovers the spanning cluster with high fidelity, and produces observables obeying the expected finite-size scaling near the critical point. We observe that to get such performance it is key that the learned latent representation exhibits critical fluctuations and scale-dependent flows consistent with the renormalization-group structure of percolation.

Categorizing continuous data into discrete bins is a fundamental operation in artificial intelligence. We introduce the categorizer automaton, a deterministic automaton that reads an infinite sequence of rewards and identifies which of finitely many bins contains its discounted sum. Categorizer automata generalize comparator automata, the special case of two bins, which have already proven useful in quantitative synthesis. Our main technical contribution is the construction of a categorizer automaton whose state space is linear in the number of bins, rather than exponential as obtained by a cross-product of comparator automata. We then apply categorizer automata to Markov decision processes, where they allow one to synthesize policies that maximize the expected utility of a discounted-sum payoff for utility functions that may be discontinuous. For piecewise-constant utility functions, the resulting algorithm is exact and runs in pseudo-polynomial time. For piecewise-Lipschitz utility functions, a class that includes any utility with bounded slope between finitely many jumps, it again runs in pseudo-polynomial time and yields an $\varepsilon$-optimal policy. We also show that the synthesis problem considered is PSPACE-hard already for piecewise-constant utilities.

Rerankers, reward models and multi-document QA scorers score candidate documents or responses in one LLM prompt, so each score depends on their order. Such scorers are selected on ranking quality, but their scores determine a decision: what a score threshold retains, a reader answers, or a preference model selects. However, equal ranking quality does not imply equal decisions: on passage reranking, five trained scorers within 0.010 nDCG@10 retain sets that overlap by only 0.66-0.84 when reordered. A published reranker takes the highest retained-set F1 in our comparison and still overlaps by only 0.667. No prompt-time change we test removes that order dependence: the only one that gains ranking quality leaves all three decisions unchanged. Order-consistency SFT (OC-SFT) attenuates it in the weights, training a candidate's score not to depend on the order. It holds ranking quality and leads every decision-stability measure among trained scorers on all three tasks: it flips the reader's answer on 0.125 of permutation pairs against 0.149-0.164 for three other objectives that target order. It is more stable than order-averaged distillation on 12 base models, and one OC-SFT permutation retains sets that overlap more than ten averaged off-the-shelf permutations. A comparison should therefore report what a threshold retains and a reader answers, not ranking quality alone. Code is available at https://github.com/thomsonreuters/presentation-dependence.

Faithful chart generation in real-world data-science workflows requires grounding visualizations in scattered evidence, computing chart-ready quantities, and rendering them accurately. Modern LLMs can produce visually plausible, instruction-compliant charts, yet data-level hallucinations remain difficult to detect in long, noisy, and multimodal contexts. To measure this gap, we introduce DEEPCHART, an expert-annotated benchmark of 1,482 task-conditioned chart-generation instances drawn from real-world scientific papers, financial filings, and ecosystem reports. DEEPCHART formulates chart generation as an Extract--Reason--Visualize pipeline and evaluates source-data extraction, derived-data reasoning, and chart rendering stage by stage. Experiments with state-of-the-art models show that visually plausible charts often conceal data-level hallucinations, with extraction and reasoning errors common in realistic long and multimodal settings. These findings suggest that larger context windows alone are insufficient; faithful chart generation also requires reliable evidence extraction and quantitative reasoning before rendering. Our benchmark and associated resources are available at https://github.com/tangdouer1005/DeepChart.

Contextual bandits are a standard framework for sequential decision-making under uncertainty, with applications in clinical trials, dosage selection, recommendation systems, and autonomous systems. Safety is central in many of these applications, since a single unsafe decision in settings such as dosage selection or autonomous driving can have catastrophic consequences. A common way to model safety in bandit problems is to associate each action with both a reward signal and a cost signal, and to optimize reward subject to constraints on cost. Most existing safety-constrained bandit models enforce safety by requiring the expected cost of each action to remain below a prescribed threshold. However, this may be insufficient in heteroscedastic settings, where the chosen action affects not only the expected reward and cost, but also the variability of the observed outcomes. We study contextual bandits with one-dimensional continuous actions and stage-wise high-probability constraints on the realized cost. We propose High-Probability Constrained UCB, an optimistic-pessimistic algorithm that explores for reward while conservatively estimating the safe action set. For linear reward and cost models, we prove a tight $\tilde{\mathcal{O}}(d\sqrt{T})$ regret bound, and we extend the analysis to general function classes using the eluder dimension. Experiments show that enforcing realized-cost safety substantially reduces violations compared with expected-cost constrained baselines.

LLM agents used for scientific experimentation must do more than generate executable code: they must implement the reference method faithfully, design experiments that test the paper's claims, and provide evidence supporting those claims. We show that agents often produce methodological hallucinations: silently reducing datasets or training budgets, replacing failed learning or generative components with lookup or oracle functions, or drawing conclusions from resource-limited settings where a method's claimed advantage disappears. To detect these failures, we introduce ABE-Ralph, a reference-anchored auditing framework that represents claims, protocols, required components, baselines, and metrics as structured experimental constraints, guides implementation through an 8-step workflow, and performs quantitative, qualitative, and code-level verification. Across 30 long-horizon reproduction runs covering 12 machine learning domains, ABE-Ralph achieves a 93% robust execution rate and identifies five scientific failure modes. In 23 NatureBench discovery tasks, ABE-Ralph matches or exceeds state-of-the-art performance on 5 tasks. These results show that reliable evaluation of AI scientists must assess whether the experimental design faithfully tests the intended claim and whether the resulting evidence supports it, rather than treating code execution or plausible metrics as evidence of scientific success.

Data lakes rely on metadata to remain usable, yet this meta data is often limited or weakly informative for column relationship discovery, especially in ERP-derived datasets with coded or abbreviated schema labels. We propose ColRel, a two-stage method that builds column embeddings from metadata and data available at ingestion time. In difficult cases, such as coded schemata, business dictionaries help better interpret column names and support the generation of short natural-language descriptions used in the second stage. Experiments on public benchmarks and an industrial ERP dataset show that ColRel is particularly effective in semantically related, weak-signal settings.

The distribution of spoken word duration of English homophones is known to co-vary with frequency of use. This study investigates whether other aspects of the phonetic realization of homophones also differ. A series of quantitative investigations of 14,000 homophone tokens in American television news broadcasts revealed that the tokens of homophone pairs such as \textit{weight} and \textit{wait} have different phonetic realizations, and that these can be predicted from their meanings in utterance context. These systematic differences remain even when taking duration-related variation into account. Time-normalized spectrograms emerged as an excellent tool for probing the fine details of phonetic realization, and obviate the need for phonetic transcriptions, which inevitably hide the phonetic truth from our eyes.

Diffusion models can be used to generate spatiotemporal signals of physical phenomena, such as time-series images of fluid dynamics. However, a major limitation of standard diffusion models is that they do not incorporate constraints derived from the underlying physical laws. Consequently, generated samples may appear visually plausible while deviating substantially from the true dynamics. In this study, we propose a simple yet effective physics-informed approach based on diffusion guidance with self-generated data augmentation. The proposed method learns the data distribution conditioned on the degree of deviation from the physically correct dynamics and generates samples by explicitly setting the deviation condition to be zero. The method decouples the evaluation of the governing equations from the diffusion model training and sampling processes, avoiding the need to solve the governing equations at every iteration of the denoising process. This design makes the method applicable to problems requiring computationally expensive numerical simulations and enables faster sample generation. Experimental results demonstrate that the proposed model not only significantly reduces the deviations compared with standard diffusion models but also achieves further reductions when combined with existing physics-constrained diffusion methods.

Scientific LLM agents have shown promise in literature reasoning, tool use, and experiment planning, but it remains unclear whether they can autonomously improve large, tightly coupled scientific machine-learning systems through executable code changes and computationally expensive validation. We study this question in protein folding, where progress requires coordinated architectural modifications, multi-objective evaluation, and domain-aware interpretation. We present AgentFold, a multi-agent framework that formulates folding-model development as a closed-loop search over executable code variants. Starting from ESMFold, AgentFold proposes hypotheses, implements and debugs code-level modifications, evaluates model variants, analyzes experimental outcomes, and stores both successful and failed interventions in structured memory. An MCTS-style policy allocates computational resources across high-scoring search branches. On an engineering-scale protein-folding codebase comprising more than 2,000 lines of code, AgentFold explores approximately 80 model variants using approximately 5,000 GPU-hours and 170 million LLM tokens. Under a matched computational budget, AgentFold improves the best lDDT by 7.5% over independent Codex proposals and outperforms a random-search control. Beyond model improvement, the resulting intervention traces reveal recurring empirical design patterns: stable gains tend to arise from early, soft, learnable priors and gated refinement, whereas direct geometric perturbations and geometry-conditioned feedback often destabilize training. The code and experimental resources are publicly available at https://github.com/lmqfly/AgentFold.

Generated operational programs are often validated with either a few hand-written examples or exhaustive regression suites. The former can miss sparse boundary and interaction faults, while the latter can be unnecessarily expensive. We introduce FaultLens, a method for learning compact behavioral test suites while preserving an auditable connection to executed evidence. It executes a rich probe domain once, stores the fault-probe kill relation as a sparse outcome cache, and learns probe orderings only from earlier program generations. A fault-driven greedy component exploits known kill structure, while a mutation-independent diversity component covers probe families, cases, templates, and temporal bins. Their alternating hybrid remains useful when a new program contains a fault mechanism absent from ordering construction. We evaluate twenty generated operational policies across four environments, ten execution seeds, 1,200 measured run summaries, 2,160 controlled program transformations, and 4,120,200 executed program-probe pairs. Of 1,960 intended faulty transformations, 1,779 alter a contract or output somewhere in the finite audit domain; 200 additional controls preserve behavior. A 32-probe hybrid learned on generations 1-3 covers 576/582 (99.0%) dynamically killable faults in generations 4-5 using 1.2-2.0% of the exhaustive domain. With an entire fault family withheld from training, diversity raises scenario-family macro coverage from 84.6% to 94.9%. In a downstream deployment study, a conservative admission rule reduces severe tail regressions from 15/20 program-environment groups to 0/20. FaultLens provides a prioritized evidence mechanism, not a proof of correctness, and makes its budget, evidence source, generalization split, and misses explicit.

Enterprises fine-tune language models on proprietary data that may later require removal due to privacy, contractual, or compliance obligations. Selective unlearning removes requested knowledge while preserving model utility, offering a practical alternative to full retraining, but existing methods treat the explicitly identified forget examples as the complete deletion scope. This is insufficient when target knowledge remains recoverable through paraphrases, aliases, or neighboring training examples. We propose GRAPHSU, a graph-guided controller that expands the deletion scope beyond forget seeds by constructing a weighted support-route graph, propagating deletion pressure through it, and applying graded forgetting strengths to high-risk neighbors. On the Task of Fictitious Unlearning (TOFU), a synthetic author-profile question-answering benchmark, and PISTOL, a structural-unlearning benchmark built around interconnected factual samples, with GPT-2 Medium and Llama-3.2-3B-Instruct, GRAPHSU achieves the lowest utility-feasible soft leakage across all deletion settings, reducing leakage by up to 49.5 percentage points over a matched seed-only baseline, demonstrating that effective enterprise unlearning requires controlling support routes, not just forget seeds.

Claude Code is an agentic work environment: a language model operating in a loop with filesystem access, shell execution, browser control, scheduled and cloud execution, external tool connections through the Model Context Protocol, and multi-agent orchestration. Its capability envelope now exceeds what one practitioner can supervise by attention alone, and its failure modes are systemic rather than local: an unreviewed hook, an over-scoped connector, a stale completion condition, an autonomous routine inheriting every credential on an account. This book is a task-oriented reference for operating that system safely and productively, written for practitioners accountable for the result. It advances four propositions. First, capability without a defined and observable completion condition is not productivity. Second, instruction, permission enforcement, sandboxing and operating-system isolation are four distinct layers of a control stack, only two of which are enforced, and conflating them is the most common cause of loss of control. Third, third-party skills, plugins, marketplaces, channels and MCP servers are software supply-chain dependencies and must be governed as such. Fourth, the correct unit of trust in agentic work is observed evidence, not an agent's closing statement. Thirty-four chapters run from installation to a fully verified capstone, with a governance part on managed policy, data residency and retention, observability and accessibility. Every product claim carries a citation to a primary source; an evidence ledger records where a claim in circulation was found wrong, what a later re-verification changed, and what remains unverified. Controls are mapped to seventeen external frameworks in a crosswalk, and an organisational adoption maturity model is proposed. Claims not confirmable from primary sources are labelled UNVERIFIED rather than softened.

Outdoor LiDAR semantic scene completion (SSC) recovers a dense semantic voxel grid from a scan observing 1% of the target volume, under class imbalance beyond 7,000x. We recast SSC as generative semantic scene completion (GSSC): a single discrete-diffusion formulation in three roles. First, paired sparse-dense scene synthesis (PS$^3$) generates matched sparse LiDAR observations with their dense semantic completions, addressing the long tail at its source and yielding the PS$^3$-SemanticKITTI corpus we train on alongside SemanticKITTI. Second, semantic-guided generative scene completion (SGSC) generates the scene from noise with multinomial discrete diffusion, conditioned on the sparse scan through a bird's-eye-view semantic map and a sparse 3D feature stream. Third, the same framework instead refines an existing completion in one flow-matching step: structured source discrete diffusion (S$^2$D$^2$). S$^2$D$^2$ improves the mIoU of SGSC's own output and every external SSC base tested, without base retraining or test-time adaptation. On the strongest base, one step without test-time augmentation reaches 38.8% mIoU on the SemanticKITTI hidden test. To our knowledge that is the best causal, single-sweep, single-sample result on that leaderboard, +2.1 pp over the previous best published score under the same restriction. Four correction steps with eight-view test-time augmentation reach 39.2%, outside that restriction.

Specializing large language models to vertical domains improves domain-specific behavior but often degrades general capabilities such as reasoning, coding, instruction following, and creative writing. We study this domain--general trade-off in Multi-Teacher On-Policy Distillation (MOPD), where a specialized student is supervised on its own sampled trajectories by domain and general teachers. Standard MOPD faces two limitations: ordinary on-policy sampling rarely exposes tokens with large positive teacher--student advantages, while the advantage sign alone does not establish whether the resulting update direction is reliable. We propose uncertainty-calibrated MOPD to address these limitations. Dual-temperature sampling broadens the candidate trajectory pool, and positive-advantage-density filtering selects trajectories with stronger positive learning signals. Centered log-likelihood (CLL) filtering then computes an entropy-calibrated teacher-endorsement score and probabilistically retains token updates according to direction--endorsement consistency. Experiments on role-playing and medical-domain specialization show that our method improves the general-capability average over standard MOPD by $4.73\%$ and $10.84\%$, respectively, while maintaining vertical-domain performance. Ablations and diagnostic analyses further confirm that the gains do not merely result from a larger rollout budget and that the proposed trajectory- and token-level mechanisms address their intended failure modes.

Agent skills bundle instructions, reference data, and executable helpers that let a general agent perform specialized tasks. Hosted providers can keep these files secret while selling access to task results, making the skill itself a valuable target. Existing disclosure defenses can block requests that ask for the skill or reproduce its text, but they cannot block customers from submitting the ordinary tasks the service is built to complete. We present Daydreaming, an execution-only attack that steals a multi-file skill through black-box task interactions. The victim is never asked to reveal the skill or grade a reconstruction. Instead, Daydreaming adaptively creates crafted tasks whose results distinguish possible hidden behaviors. It tests individual behaviors, uses attacker-controlled shadow agents to choose a design, and completes each file using stored victim results and local execution checks. We formalize three nested threat levels of access as Differential, Trace, and Output, and focus on Output, where the attacker sees only the final response and returned files. Across 7 skills and 4 victim models, Daydreaming recovers 86.8% of the original skill's capability at Output, outperforming SigLeak by almost 4x. It produces installable skills using a median of 32 victim calls per skill even with disclosure defenses enabled. These results show that hiding skill files and filtering direct disclosure do not, by themselves, prevent functional reconstruction through normal use.

Graph neural networks (GNNs) are typically conceptualized as message-passing neural networks, yet it remains unclear why neighborhood aggregation reliably outperforms node-wise multilayer perceptrons (MLPs). Despite its empirical success, this paradigm can be computationally expensive and sensitive to imperfect graph structures. In this work, we present a retrieval-augmented view of GNNs: each layer makes predictions by applying an MLP to a node representation together with a permutation-invariant summary of retrieved graph context. Motivated by this perspective, we propose RTA, a simple MLP-based framework that replaces structural message passing with label-aware retrieval and propagation. We provide theoretical insights that (i) connect retrieval-based aggregation to softmax-attention message passing, and (ii) establish the robustness of retrieved-context supervision to mis-retrieved outliers. Experiments on multiple text-attributed graph benchmarks show that RTA matches or even outperforms strong GNN and graph LLM baselines while improving efficiency and robustness across diverse scenarios.

Large language models offer broad capabilities, but adapting them to evolving domains, tools, and requirements often entails repeated post-training. Autonomous systems automate parts of this process by proposing updates, training candidates, and using evaluation feedback to select subsequent proposals. As evidence accumulates, a central problem emerges: which past update evidence remains actionable after subsequent training has changed the parent model? An update's effect depends on its parent, data, and training stage. Treating past success as context-free permission can waste compute. If the resulting child is promoted, it can also degrade the subsequent training trajectory. We formulate this problem as conditional experience transfer and introduce Boundary-Calibrated Intervention Transfer (BCIT), a method that authorizes experience reuse before weight-changing training. BCIT binds an observed effect to its source context, checks applicability conditions, vetoes candidates with named hard conflicts, and obtains current-state evidence through a bounded training trial when needed. Fully trained candidates still face a shared adoption rule, and only observed events extend memory. On one 4B model adapted across finance reasoning, text-to-SQL, and function calling, candidate updates exhibit heterogeneous target and retention effects across the evaluated contexts. Under matched candidates, evidence, and compute, BCIT authorizes fewer harmful updates and attains higher equal-budget final-model quality than the evaluated alternatives. These results support treating experience authorization as a distinct problem in autonomous post-training.

In recent years, transductive knowledge graph embedding models have been applied to tasks such as link prediction and query answering. Although knowledge graphs often contain rich numerical attributes, most embedding models neglect them, limiting their ability to represent real-world knowledge graphs with diverse information. In this work, we propose a neural regression model (LitEm) that enables transductive knowledge graph embedding models to predict numerical attributes within knowledge graphs. Experimental results demonstrate that LitEm achieves the best or second-best results on most attributes across FB15K-237, YAGO15K, DB15K, and Mutagenesis. Furthermore, we propose a co-training framework that jointly trains state-of-the-art transductive knowledge graph embedding models with LitEm, which improves link prediction performance mainly for bilinear models and simultaneously enables them to predict numerical attributes. In addition, the literal-awareness evaluation demonstrates that co-training helps models to encode and exploit attribute information in a "literal-aware'' manner, suggesting that the observed gains are not merely due to additional parameters. We publicly release our implementation at https://github.com/dice-group/dice-embeddings.

Diffusion-based Video Virtual Try-On (VVT) achieves high visual fidelity through bidirectional spatio-temporal modeling, but complete-clip dependence incurs prohibitive latency and computational overhead in practical continuous deployment. Naively enforcing causality disrupts pretrained bidirectional priors and substantially degrades synthesis quality. We introduce LiveVVT, a rolling streaming diffusion framework that preserves bounded bidirectional modeling within causal recurrent generation. Within a fixed-size window, LiveVVT jointly denoises multiple video chunks under bounded look-ahead, preserving local bidirectional interactions while emitting one clean chunk per iteration. Beyond the window, two complementary memories sustain long-term consistency: a bounded temporal memory propagates recent dynamics and occlusion context, whereas a persistent global appearance memory, constructed once from the target garment and a frontal try-on keyframe, anchors garment details and dressed appearance throughout the stream. We further introduce a progressive distillation framework integrating bidirectional VVT learning, teacher-trajectory regression for causal few-step adaptation, and Collaborative Matching Distillation, which couples teacher-distribution matching with rolling flow matching on real videos to align optimization with recurrent inference. Experiments on paired and unpaired long-sequence benchmarks demonstrate superior generation quality over similarly sized models, with $26\times$ lower latency and $11\times$ higher throughput, enabling high-fidelity real-time streaming VVT.

Recent advances in Multimodal Large Language Models (MLLMs) have extended Image Aesthetic Assessment (IAA) beyond scalar scores toward interpretable critique and guidance. Yet existing benchmarks mainly assess intrinsic visual quality or fixed domain criteria, leaving open whether an appealing image is appropriate for a specific purpose, audience, cultural setting, or domain convention. We introduce AesCanvas, a unified suite with two complementary components: CritiqueCanvas with 519,136 instruction-response pairs from 54,300 images supports long-form, multi-dimensional critique across photography, painting, and virtual imagery, whereas ContextCanvas with 301 expert-reviewed use scenarios evaluates contextual aesthetic suitability in realistic use scenarios. Under a unified protocol, we evaluate closed-source frontier, open-weight general, and aesthetic-specific MLLMs. Results reveal a clear separation between critique generation and context-sensitive judgment: reference-based lexical and semantic metrics only partially capture critique quality, while aesthetic specialists remain competitive on selected critique metrics yet substantially lag strong general-purpose MLLMs on ContextCanvas. Further analyses show that aesthetic specialization does not reliably transfer to contextual suitability and that model decisions may fail to track or ground themselves in decisive contextual visual cues. These findings establish culturally situated, evidence-grounded suitability as a distinct objective for aesthetic modeling.

AI text detectors are increasingly employed in academic settings, but it remains unclear whether their outputs reflect AI authorship itself or broader linguistic features associated with polished academic English. Previous studies have reported high false-positive rates (FPRs) for non-native English writing, but population-level comparisons confound authorship with differences in topic, domain, and writing style. Professional editing provides a useful setting for examining this issue because it changes the linguistic form of manuscripts while preserving authorship and content. We examined 135,389 document pairs from a professional English editing service (2018-2025), comprising non-native manuscripts and their native-edited versions, to assess how editing affects detector responses controlling for content and authorship. For the 13 AI text detectors, FPRs for human-written texts varied widely, from 0.0% to 100.0%. Responses varied across detectors: the same edits increased AI scores in some detectors but decreased them in others. Notably, score changes correlated with the extent of editing. The findings identify professional editing style as a key confounding variable in AI detector outputs, rather than establishing a full separation of text origin from linguistic style, raising concerns about fairness and reliability in academic settings.

Expert-level financial question answering requires both grounded verification to catch numeric hallucinations and audit trails for regulatory compliance, attributes that standard single-pass RAG systems lack. We take a step toward this goal with Self-Improving RAG, a framework that decomposes document QA into three specialized agents (Retrieval, Reasoning, and Judge) coordinated by an orchestrator with feedback-driven self-correction. When the Judge Agent scores an answer below a dynamic threshold, the system triggers retry with escalated strategies: broader retrieval, more careful prompting, and relaxed acceptance criteria. We evaluate on FinanceBench (SEC filing QA), where Self-Improving RAG achieves 86% oracle-guided accuracy (measuring agreement with gold answers) with a 36.4% Lazarus Rate, recovering nearly 4 in 10 initially incorrect answers through targeted retry. A key finding is that a fixed retrieval pipeline with judge-driven retry achieves strong results without dynamic routing, providing full interpretability. Every decision is logged with confidence scores, enabling the audit trails required for regulated financial applications.

We present an extension and comprehensive real-world validation of Co-Scientist, a Gemini-based multi-agent system designed to accelerate end-to-end scientific research across hypothesis generation, experimentation, and manuscript generation. Moving beyond in silico hypothesis generation, this specialized configuration transitions Co-Scientist into an execution-grounded research partner advancing closed-loop scientific workflows across materials science, biology, and computer science. In materials science, Co-Scientist interfaced with a semi-automated chemical vapor deposition reactor to design a safe precursor route for MXenes; experimental execution produced a lamellar 2D material sharing key structural similarities with the Ti3C2Tx MXene lattice, although further experiments are needed to confirm the atomic structure. Leveraging Gemini 3 Deep Think for rapid, lab-in-the-loop execution, it also tailored growth recipes to laboratory constraints in minutes, enabling single-attempt growth of monolayer MoS2, MoSe2, and WS2 semiconductors. In biology, Co-Scientist predicted emergent swarming phenotypes of engineered E. coli across inducer (IPTG) gradients from sparse imaging data, quantitatively matching unpublished wet-lab morphological measurements. In computer science, Co-Scientist autonomously discovered an inference-time scaling architecture that outperformed six frontier models on HealthBench (Hard and Professional) while reducing potential clinical harm under blinded physician evaluation. Finally, a double-blind study of end-to-end generated papers with 30 domain experts across 450 reviews demonstrates that Co-Scientist's reliability modules reduce hallucination and plagiarism while improving research safety. Together, these results demonstrate progress toward closed-loop multi-agent scientific AI systems capable of accelerating real-world scientific discovery.

Reply assistance in multilingual settings requires linguistic competence and culturally situated judgments of appropriateness. We present PragAlign, which separates context reading from selective clarification, and evaluate it alongside Direct and Rule. Nine native Chinese speakers judged Chinese materials, while three native Japanese speakers judged matched Japanese versions. In the Chinese evaluation, PragAlign received significantly better ranks than both baselines. In the Japanese evaluation, Direct had the lowest mean rank, PragAlign had the highest top-rank rate, and the omnibus difference was not significant. The groups selected the same top condition in 5 of 10 scenarios, including four shared PragAlign selections. The results identify shared and language-specific judgment patterns and inform reply assistance designed to support linguistic and cultural understanding.

As the most widely used container orchestration platform, Kubernetes provides flexible privilege configuration by allowing developers to manage Linux capabilities via manifest files. However, developers rely on default settings or coarse-grained security contexts in practice, violating the principle of least privilege and enlarging the attack surface of containerized workloads. Existing studies either detect vulnerable patterns in Kubernetes manifests or infer required capabilities for standalone Linux programs, but they do not directly address capability minimization in Kubernetes. To bridge this gap, we first conduct an empirical study on three open-source datasets, revealing that 74.67% of projects lack capability configurations. Motivated by our observations, we propose KubeCap, a framework for Kubernetes capability minimization. KubeCap translates deployment specifications into deterministic manifests, locates container entrypoints, performs reachability-guided system call analysis, and leverages LLM-assisted rule specification to derive syscall--parameter--capability relations from Linux kernel code. Based on these results, KubeCap infers the minimal capability set required by each workload and automatically generates repaired manifests. Evaluation on 10 representative Go-based Kubernetes projects shows an average capability reduction rate of 54.97%, outperforming rapid type analysis and class hierarchy analysis baselines while maintaining practical analysis cost. These results demonstrate KubeCap's effectiveness in enforcing least privilege in Kubernetes.

Synthetic speech provides scalable supervision for automatic speech recognition (ASR), but its benefit depends on the selected texts, reference speech, and amount of synthesized data. We present a unified phoneme-based TTS-to-ASR augmentation pipeline built around a multilingual TTS model trained from scratch using the F5-TTS architecture with language-ID conditioning. The pipeline combines language-specific grapheme-to-phoneme conversion, reference-speech filtering, candidate-text selection, synthesis, and matched ASR continuation. We further propose phoneme-frequency-guided selection (PFGS), which ranks candidate sentences using phoneme frequencies estimated from real ASR training labels. Experiments with separate monolingual ASR systems for Arabic, French, Italian, and Portuguese span 13 test sets. Across the synthesis-scale sweep, random augmentation improves over matched real-only continuation on 11 test sets. Under a nominal 60% synthesis budget, PFGS improves over real-only training on 12 test sets and over random selection on 9. Its largest relative word error rate (WER) reduction against random selection is 19.3%. With target texts and synthesis counts fixed, reference-speech filtering reduces absolute WER by 0.29 and 0.59 points on Italian and French Common Voice, respectively. These results identify synthesis scale, candidate-text content, and reference quality as important control variables in TTS-based ASR augmentation.

Enterprise deployments of autonomous AI agents inherit a control model built for human users and long-lived services, and the fit fails in three specific ways: agent principals are ephemeral, appearing and vanishing faster than provisioning; their actions are selected by a model rather than programmed, so the set of things they may attempt is not known in advance; and the population is discovered rather than provisioned, because anyone who can call an API can create one. We argue that governing such agents is a runtime problem -- not a model-alignment problem and not a build-time problem -- and we derive five primitives from the questions that must be answered before an action takes effect and after it has: discovery, identity, governance, attestation, and supply chain. For each we state what fails if it is absent and why the others cannot structurally supply it. We describe an implementation in which an agent's action is mediated against policy before it takes effect, authorised against a per-tenant action vocabulary, and recorded in a hash-linked signed ledger a third party can verify with the vendor out of the loop. We report what the architecture costs: the enforcement point sits on the request's critical path, identity requires a sidecar per workload, and fail-closed mediation converts availability incidents into denial. We are explicit about implementation status: four primitives are built and running in private pilots, and the fifth is built as separate tooling and not yet integrated into the request path. We keep it in the set deliberately: a five-part decomposition that exactly matches what its authors happened to build is not a taxonomy but a description of a codebase.

Detecting AI-generated text (AIGT) remains challenging because existing approaches rely on token-level statistical signals or independent stylometric features, causing them to overfit to specific generators and fail under distribution shift. We identify a structural signal at the sentence-pair level: LLMs produce inter-sentence transition variance that deviates from human writing through inflated variance driven by recurring similarity bursts at paragraph boundaries and templated transitions. We formalize this as Relational Over-Regularization (ROR) and validate it across four benchmarks (p < 0.001). The central contribution is this relational problem formulation, not a novel GNN architecture; CSFG is one concrete instantiation for operationalizing ROR. To exploit this signal, we propose the Cross-Source Stylometric Fingerprint Graph (CSFG), a graph-based framework that encodes positional, sequential, semantic, and transition deviation signals as learnable GNN edge features. The per-edge signed deviation δ_ij operationalizes ROR without hand-crafted thresholds and acts as a false-positive calibrator. CSFG achieves 97.14% accuracy under binary detection, outperforming the strongest graph-based baseline by 11.14 pp, with a false-positive rate of 1.57% and robust generalization to unseen LLMs in the inflated-variance regime; detection degrades for generators whose transition variance falls at or below the human baseline.

While AI-assisted text-based counseling is gaining attention, it remains empirically unclear which counselor behaviors are associated with higher dialogue quality. Existing research often focuses heavily on Reflection, borrowing frameworks from Motivational Interviewing. To address this gap, we conduct a multi-layered analysis using KokoroChat, a large-scale Japanese text counseling dataset conducted by professional counselors and trainees, newly annotated with counselor strategy tags and client distress levels. Our results show that, under the quality indicators used in this study, Affirmation is more consistently associated with session quality than Reflection among the analyzed strategies. Cross-dataset transfer experiments further suggest that this quality signal can be observed to some extent on ESConv, an English dataset with non-expert supporters. These findings provide empirical implications for counselor training and emotional support system design. We release the additional KokoroChat annotations and experimental source code at https://github.com/UEC-InabaLab/BeyondReflection.

Despite the growing number of public datasets, annotated medical images remain scarce. Supervised learning methods achieve strong performance on many benchmarks, however require large amounts of labeled data, which are costly and time-consuming to obtain in the medical domain. To address this limitation, contrastive self-supervised learning (SSL) has emerged as a promising alternative for learning useful representations from unlabeled data. In this work, we investigate two SSL frameworks, SimSiam and SimCLR, for retinal disease classification from fundus images. We focus on understanding how augmentation strategies and training parameters influence representation learning under resource-constrained settings. Given limited data and computational capacity, we explore the feasibility of training SSL models with small batch sizes incorporated with retinal-specific augmentation techniques. Through a series of experiments, we assess the quality of learned representations via linear evaluation and fine-tuning across downstream tasks, including multi-disease classification and diabetic retinopathy grading. Our results show that tailoring augmentation strategies to the characteristics of retinal images plays a critical role in improving performance. Even under constrained settings, lightweight SSL frameworks can learn transferable representations that reduce dependence on large annotated datasets and achieve competitive results.

Identifying white space --- the unexplored but potentially valuable regions of a patent landscape --- is essential for strategic R&D planning, yet existing methods rely on manual patent mapping or apply single-view clustering without quantitative gap detection. We propose BLANC (Blank Landscape Analysis through NPMI Conditioning), a three-phase pipeline combining (1) multi-view neural topic modeling along three semantic dimensions (application/use, novelty, inventive step); (2) Normalized Pointwise Mutual Information (NPMI) to quantify cross-dimensional cluster association; and (3) conditional detection that flags combinations whose NPMI drops when the corpus is filtered by a user-specified keyword. The drop is captured by a new metric, $Δ$NPMI, which identifies combinations "established globally, unexplored locally." Because white space has no ground truth, we evaluate BLANC on two public USPTO corpora --- machine learning/AI (5,417 patents, CPC G06N) and glass compositions (1,982 patents, CPC C03C) --- by artificially depleting known technology combinations and testing recovery. When three-quarters of a target pair's documents are removed, BLANC recovers 34.1% (ML/AI) and 27.3% (glass) of the depleted combinations, whereas size-matched removals not aimed at them (random documents, or those of a different established combination) essentially never do: the target is never recovered in 191 decoy trials. Collapsing the three semantic views into one recovers nothing, while prior co-occurrence measures also flag the target under random removal, offering no specificity. In a proprietary case (302 float glass / glass-ceramics patents), the keyword "fluorine" reveals a fluorine surface treatment $\times$ warpage suppression candidate ($Δ$NPMI up to 0.48) that experts had independently identified.

Cooperative multi-agent reinforcement learning (MARL) faces significant challenges in maintaining robust coordination under noisy observations. Although observation disturbances are often introduced independently across agents, their downstream effects on cooperative decision-making can become structured through underlying cooperation structures. We characterize this phenomenon as structured noise effects, where noise-induced decision effects exhibit local correlation among agents with stronger task-related dependencies while remaining globally heterogeneous across different agents and local structures. Existing robust MARL methods, however, rarely explicitly characterize or exploit such structure-dependent noise effects. To address this limitation, we propose SIGMA, a hierarchical collaboration framework that exploits cooperation structures to learn robust representations under noisy observations. SIGMA first organizes agents into adaptive local structures through density-based grouping and performs intra-group consensus aggregation to preserve shared task-relevant information while smoothing agent-specific representation deviations. Inter-group attention then adaptively integrates information across different groups to preserve global coordination while accommodating their heterogeneous contributions. Experiments on noisy-observation tasks in StarCraft II empirically validate the structured noise effects and demonstrate that SIGMA consistently improves robustness under observation noise while maintaining competitive performance in noise-free environments.

Pruning is a practical approach to compress large language models (LLMs), but it can amplify text degeneration, especially repetition loops, even when perplexity and task accuracy remain largely unchanged. In this work, we present a token-level analysis of this failure mode by viewing decoding as a dynamical process that enters and persists in a small set of recurrent contexts. Our analysis decomposes degeneration into loop entry risk and loop persistence, and shows that persistence is controlled by the escape mass assigned to plausible alternatives within the token sampling set. Motivated by these findings, we propose two token-level guidance objectives for post-pruning fine-tuning. FOCUS reweights distillation toward high-confidence teacher regions to suppress leakage, while RePAIR uses onset-centered positive/negative continuation pairs with a margin loss to promote plausible alternatives and prevent early commitment to repetition loops. Experiments on open-ended continuation and instruction-based generation show that both methods consistently reduce repetition and improve generation quality.

As large language models are increasingly deployed to simulate diverse human characters, ensuring persona fidelity, defined as the extent to which an agent's behavior consistently reflects the psychological and stylistic characteristics of a target persona, has become a critical requirement. However, existing evaluation paradigms primarily rely on either holistic LLM-based judges, which are prone to "holistic appraisal hallucination'', or static psychometric inventories, which fail to capture the context-dependent fidelity required in dynamic dialogue. To address these limitations, we propose PRISM (Persona Reasoning with Inverse SFL-based Modeling), a psycholinguistically grounded framework that reformulates persona fidelity evaluation as a structured inverse inference task. Inspired by Systemic Functional Linguistics (SFL), PRISM decomposes persona fidelity into three functional dimensions: Task Framing, Interpersonal Stance, and Linguistic Style. It estimates dimension-specific evidence over a persona-conditioned label space and aggregates these signals into an interpretable and auditable evaluation process. Experiments show that PRISM yields more accurate and stable judgements than traditional holistic judging, providing a more reliable framework for persona fidelity evaluation.

Recent progress in offline reinforcement learning (RL) has been driven by expressive generative actors such as diffusion and flow-matching policies, which capture multimodal behavior in offline datasets. However, these actors require multiple denoising or integration steps per action and thus incur substantial overhead at every decision in deployment. In this work, we revisit where capacity should be invested in an offline actor--critic method. Since the critic is used only during training and is discarded at deployment while the actor runs at every decision step, allocating capacity to the critic rather than the actor is more favorable for inference-time efficiency. However, scaling MLP critics in offline RL is known to introduce several distinct instabilities that have, in practice, kept critics shallow. We identify three distinct failure modes that arise when critics are deepened in offline RL---optimization, bootstrap-noise amplification, and value-range drift---and address each with a corresponding ingredient: a residual MLP backbone, n-step bootstrap targets, and a categorical cross-entropy loss. Combining these ingredients with a lightweight deterministic actor, we propose LAC (Light Actor, deep Critic). On OGBench, LAC matches the strongest diffusion- and flow-matching baselines while achieving up to 4x lower inference latency, comparable to one-step distilled policies without distillation. Its critic recipe also transfers across actor parametrizations.

Generative retrieval has demonstrated strong performance by directly generating product semantic identifiers (SIDs). Extending this paradigm to image search, however, is nontrivial because real-world query images contain diverse information, including the search target, useful auxiliary evidence, and irrelevant visual content. This requires the model to identify and focus on the search target while selectively utilizing auxiliary evidence. In this paper, we propose \textbf{PailitaoGR}, a \emph{Latent Think-with-Images} method for generative image retrieval, which internalizes target-focused perception and selective auxiliary-evidence utilization into a the generative retrieval model, enabling \textit{Zooming without Cropping} and \textit{Reading without OCR}. Specifically, we design a target-focused perception mechanism that identifies and enhances visual tokens of the search target, consisting of a target Enhancer and a learning strategy based on on-policy distillation and attention guidance loss, enabling the model to focus on search-target regions. We also design a selective auxiliary-evidence utilization mechanism that identifies and enhances visual tokens of auxiliary evidence, including an auxiliary enhancer and an in-capacity incremental contrastive distillation strategy, enabling the model to exploit auxiliary evidence. We construct training and validation sets sampled from real-world online image-search logs. Experiments show that our method outperforms existing baselines by an average of 13.8\%, validating its effectiveness.

Multi-object removal in 3D scenes is challenging due to severe occlusions, semantic entanglement, and the difficulty of maintaining geometric and multi-view consistency. Existing 3D Gaussian Splatting (3DGS) methods perform well for single-object editing but scale poorly to multi-object scenarios, often requiring repetitive optimization and yielding unstable geometry in removed regions. We propose CoGeo-GS, a concept-driven framework for controllable multi-object removal in 3D scenes. CoGeo-GS assigns concept-aware semantic tags to Gaussians, enabling flexible object selection and reducing interference between foreground objects and background structures within a single optimization stage. To recover plausible geometry, we introduce a geometry-aware completion pipeline that combines monocular depth priors with diffusion-based refinement and boundary-aligned blending. A geometry-regularized refinement strategy further stabilizes reconstruction and preserves multi-view consistency. Experiments demonstrate that CoGeo-GS outperforms existing methods in visual quality and reconstruction fidelity.

Model merging promises to construct a single multi-task model from independently fine-tuned task models without accessing the original task data. This makes it attractive when task data cannot be centralized, but released task models may still leak private fine-tuning data. Differential privacy (DP) provides a principled mechanism for limiting such leakage, yet its effect on model merging remains poorly understood. In this paper, we study the geometry of differentially private model merging and identify two geometric obstacles that make private task models difficult to merge: \emph{local sharpness}, which makes task losses sensitive to the parameter displacement induced by merging, and \emph{reference drift}, which measures the displacement of private task models from the shared pretrained initialization and amplifies cross-task interference. Based on these observations, we propose \textbf{DP-Merging}, a geometry-aware framework that improves the mergeability of differentially private task models. DP-Merging uses a DP-compatible sharpness-aware objective to guide each private task model toward flatter loss regions, and a reference-based alignment regularizer to keep task models close to the shared pretrained initialization. We derive a merge-gap upper bound showing that reducing local curvature and reference drift tightens the bound on the loss increase induced by merging. Experiments on vision and language tasks across multiple privacy budgets show that DP-Merging consistently improves private merged-model performance while preserving the privacy guarantees of the underlying DP fine-tuning procedures.

Mixture-of-Experts (MoE) models route each token to a subset of expert networks, increasing capacity while keeping per-token computation sparse. In many deployed MoEs, the number of active experts is fixed across layers and tasks, although layer roles and expert redundancy vary with depth and demand varies with difficulty. Existing approaches address only part of this setting: layer-wise allocations are usually determined offline and reused for all tasks, while token-level methods vary expert activation using local routing signals without task-level context. We propose MetaNet, a support-set controller that predicts, for each layer, an expert-retention threshold and a bounded routing bias. The backbone, experts, and router remain frozen. On DeepSeek-MoE-16B-Chat, MetaNet provides a tunable accuracy-expert-activation trade-off. Relative to fixed k=6, a conservative setting activates 3.61 experts on average (40% fewer) and achieves comparable MMLU accuracy (0.489 vs. 0.474), whereas an aggressive setting activates 2.28 experts on average (62% fewer) with accuracy approximately 3.7 percentage points lower. The MMLU-trained controller also transfers to C-Eval without retraining, activating 2.90 experts on average (52% fewer than fixed k=6) at 0.386 accuracy.

Objective: Model-based closed-loop neural stimulation holds promise for therapeutic applications ranging from Parkinson's disease to sensory restoration, but deployment has been limited by two obstacles: 1) forecasting models for predicting the consequences of stimulation fail catastrophically on a meaningful fraction of sessions, and 2) per-session calibration requirements are often incompatible with clinical constraints. We address both by demonstrating, for the first time, that meta-learning and pretraining can be applied to neural stimulation response modeling. Methods: Temporal basis function models (TBFMs) forecast state-dependent neural responses to stimulation. We extend TBFMs with cross-session pretraining using a novel architecture and algorithm based on model-agnostic meta-learning (MAML), evaluating them on 40 sessions of optogenetic stimulation in primary sensorimotor cortex of two non-human primates. Results: Meta-learning substantially reduces catastrophic forecast failure: for a 1k calibration set size, sessions with test R-squared < 0.05 drop from 16 of 40 (single-session training) to 1 (MAML-pretrained), and prediction intervals become significantly narrower (p < 0.05). Calibration requirements are reduced by 50-90% at matched accuracy, enabling experiments otherwise infeasible within clinical session-time constraints. Conclusion: Our results demonstrate that cross-session structure in stimulation responses is consistent enough to support pretraining, providing the first empirical evidence that meta-learning approaches are viable for neural stimulation. Significance: The robustness and sample efficiency gains directly address known obstacles to deploying model-based stimulation controllers. Our results motivate community efforts to assemble standardized multi-site stimulation datasets and to further explore meta-learning for robust closed-loop stimulation.

Many synthetic-image detectors produce accurate predictions but offer limited insight into how those decisions are formed. This paper introduces Hierarchical Channel Stacking (HCS), a compact framework for AI-generated image detection that converts intermediate CNN activations into a structured 60-dimensional representation organized across three progressively deeper backbone stages. HCS uses per-channel Level-1 classifiers and a Level-2 aggregator to produce image-level predictions while preserving explicit hierarchical structure for analysis. On a benchmark spanning GAN and diffusion generators, HCS achieves 86.7% accuracy and 86.7% macro-F1 on the held-out test set. Stage ablation shows that the full three-stage system outperforms reduced single-stage and two-stage variants, indicating that the hierarchy carries complementary predictive information. Stage-level contribution analysis further shows that, in the analyzed detector setting, fake GAN and fake diffusion images exhibit distinct stage-level contribution profiles. These results position HCS not simply as a compact detector, but as a structured framework for studying how synthetic-image detectors assemble evidence across representation levels.

Decoding quality in diffusion multimodal language models (dMLLMs) depends heavily on the order in which masked tokens are committed. Existing confidence-based strategies prioritize locally easy tokens, but confidence does not necessarily reflect contextual usefulness. As a result, structurally easy tokens such as punctuation may be committed before informative semantic anchors, weakening context propagation and increasing error accumulation. We propose Information-Guided Frontier Decoding (IGFD), a training-free decoding strategy that ranks candidates using token confidence, neighborhood uncertainty, and structural commitment risk. IGFD encourages early commitment of reliable semantic anchors while delaying fragile structural tokens, improving contextual support during decoding. A dynamic candidate frontier further constrains token selection to locally expandable regions under the same decoding budget. The method requires no additional training, auxiliary models, or extra forward passes. Experiments across multimodal understanding, reasoning, grounding, and hallucination benchmarks show that IGFD consistently outperforms existing decoding strategies across the majority of benchmarks and diffusion MLLM backbones under identical decoding budgets.

Across various non-verifiable tasks, human evaluation is reliable but expensive, while automatic metrics are more scalable but often biased. Building on prediction-powered inference (PPI), we propose prediction-powered evaluation, a framework that combines limited human judgments with large-scale automatic scores to obtain data-efficient system comparisons that are provably unbiased. We develop parametric and non-parametric procedures, analyze the efficiency trade-off between paired and unpaired designs, and validate the framework on six WMT datasets. We further introduce the Prediction-Powered Saving Ratio (PPSR), a meta-metric that measures how much human annotation an automatic metric can save when used within prediction-powered evaluation. PPSR directly targets metric utility for prediction-powered evaluation and yields more discriminative and stable metric rankings than existing system-level meta-metrics. Overall, our new paradigm reframes automatic metrics as tools for reducing human annotation cost rather than replacing human judgment, and applies broadly to non-verifiable tasks.

This report presents a framework to help organisations, policymakers and researchers reason about the risks that emerge when AI agents interact with each other, how those risks change as interactions cross organisational boundaries, and the controls that may help address them. As organisations deploy AI agents, those agents will increasingly interact with each other: inside the organisation, with the agents of partners, customers and suppliers, and with unknown counterparties on the open internet. Failures can emerge from the interactions themselves, and once those interactions cross an organisation's perimeter, no single organisation can fully see, control or govern them. The report introduces three deployment tiers, defined by the minimum common governance binding any two interacting agents: singular governance, where one organisation governs every agent; federated governance, where multiple organisations deploy into a shared environment under agreed rules; and open environments, where agents operate with no central authority and shared standards are adopted voluntarily if at all. Within each tier, the report examines risk factors, failure modes and available controls. It identifies who is positioned to apply the controls, and where no actor is positioned to act, it characterises the gap and the collective action required to close it.

LLM judges are widely used to evaluate agentic tool-calling systems, yet their reliability on structured, dependency-driven workflows remains largely unexamined. We present AgentJudgeBench, the first benchmark to systematically study LLM-as-a-judge reliability for agentic tool-calling over workflow DAGs, as distinct from the broader LLM-as-a-judge task of open-ended text or preference evaluation. The benchmark comprises 3,808 instances spanning six DAG topologies and three difficulty tiers, evaluated with five generators (3B-70B open-weight models and GPT-5.4) and six judges (20B to frontier scale) under paired with- and without-ground-truth conditions. Judge alignment degrades monotonically with task difficulty, 1.5x faster without ground truth, and on hard queries without ground truth all six judges converge to a narrow 77-82% band regardless of scale, revealing a structural ceiling driven primarily by task difficulty, though its height is partly prompt-dependent for weaker generators, that model capacity alone cannot overcome. Ground-truth exposure is not uniformly beneficial: it reduces alignment for GPT-5.4 (1.5 pp) and Gemini-2.5-Pro (3.9 pp), consistent with over-anchoring. Among mitigation strategies, chain-of-thought reasoning and judge temperature both have negligible effect, while structured evaluation rubrics improve alignment by up to 6.5 pp but do not generalize uniformly across judge-generator pairs. With ground truth, QwQ-32B best matches the programmatic reference, while a human validation study identifies GPT-OSS-120B as the most human-aligned judge; without it, frontier judges lead only marginally within the shared ceiling. These results expose fundamental limitations of current LLM judges and yield practical guidelines for reliable evaluation in agentic systems.

Model selection becomes particularly challenging under strong predictor dependence and model-class uncertainty, especially when there are exponentially many models. We propose a Descriptive-Complexity Information Criterion (DCIC) that regularizes large candidate model collections through Kraft-admissible code lengths. Under sub-Weibull noise, we establish selection consistency through approximation-error separation without relying on RIP-type conditions, together with nonasymptotic oracle risk bounds that remain valid under model misspecification. The same coding principle places heterogeneous classes on a common complexity scale at a small additional class-identification cost. This extension yields class--model recovery under suitable identifiability conditions and risk adaptation across classes. We further develop a complexity-guided search path that makes the computation--statistics trade-off explicit. Large penalties yield polynomial-size retained search regions with high probability, whereas smaller penalties sharpen the oracle risk benchmark. Numerical experiments illustrate stable support recovery and favorable estimation performance under strong dependence and model-class uncertainty.

Processing/p5 libraries across different programming languages enact consistent priorities for creative coding as a designed experience. While different programming language ecosystems, like Java and JavaScript, are each associated with their own affordances, community norms, and patterns of use, Processing/p5 sketches across these languages share similarities. Based on case studies of building an implementation of Processing/p5 in two host languages, JavaScript and Lua, we propose a list of software decision-making guiding aspects that constitute Processing/p5, regardless of host language. We discuss this framework in the context of decisions in other exploratory and creative tools that demonstrate how each of the guiding aspects can be operationalized differently than in the case studies. The proposed list highlights opportunities for learning, research, and artistic practice through creation of new Processing/p5 libraries for creative coding and algorithmic art.

This study provides a comprehensive benchmarking of conventional machine learning (ML), ensemble learning, deep neural networks, recurrent architectures, Transformers, graph based models, and hybrid ensemble deep learning approaches under complementary renewable energy scenarios. Three datasets are considered: a large scale WEC dataset, a 16 WEC dataset, and operational 10 min SCADA measurements at the Penmanshiel wind farm. For structured WEC layout data, tree ensembles exhibited a clear advantage over conventional ML and neural predictors because randomized partitioning and boosting efficiently captured nonlinear layout power interactions without requiring explicit feature representation learning. The Extra Trees was the strongest model, achieving considerable results. Relative to the MLP baseline, this corresponds to an approximately 63.7% reduction in MAE, demonstrating the suitability of randomized tree ensembles for high dimensional structured WEC data. Also, STGCN reduced the MAE to approximately 167.0 kW and achieved R = 0.93 by explicitly learning spatial and temporal turbine interactions. The best overall forecasting accuracy was obtained by the RF BiLSTM hybrid, with an MAE=150.5 kW. Compared with standalone LSTM, this represents an approximately 75% reduction in MAE, while improving on STGCN by approximately 10.0%. Finally, the experiments reveal that no single AI architecture is universally optimal: randomized and boosted ensembles are particularly effective for structured WEC surrogate modeling, graph networks become advantageous when explicit spatial interactions dominate, and ensemble recurrent hybrids provide the strongest balance when nonlinear tabular relationships and temporal dynamics coexist.

In university academic advising, identical questions can require different answers depending on a student's department, admission cohort, and degree program, causing profile-blind retrievers to surface plausible but inapplicable evidence. We present proFILL, a method for transforming hoBIT, our college's current rule-based advising chatbot, into a profile-aware retrieval-augmented generation (RAG) system. Rather than requiring a complete user profile upfront, proFILL progressively acquires only the profile attributes needed for each query, guided by both the query intent and the initially retrieved evidence, and uses them to condition retrieval over a profile-aware index. Extensive experiments and a human preference study show that proFILL outperforms diverse RAG baselines, is preferred by target users, and remains effective with open-weight models for cost-effective on-premise deployment.

Large software repositories are often beyond model context limits. Training repository knowledge into models is costly and quickly stale, while local retrieval can miss scattered requirements, and explicit relation graphs add ongoing maintenance burden. We propose an entity-only external interface with task-conditioned relation materialization during inference. A two-layer index separates global routing from local entity focus and is evaluated on DeepSeek-V4-Flash and SWE-bench Verified. The base, one-layer, and two-layer conditions achieve 92.1%, 94.2%, and 95.6% success, respectively, under zero pre-built entity-relation edges.

Multimodal large language models (MLLMs) are increasingly capable scientific assistants, yet they remain far from fully autonomous research. This transition requires models to actively inspect academic papers, build global evidence views, and make traceable judgments without prespecified issues or evidence. However, existing work provides limited task paradigms or training studies for such issue- and evidence-absent verification. We study this challenge through scientific error detection, where models must determine whether errors exist and justify them with evidence-based reasoning. To fill this gap, we present VERA-RL, a reinforcement-learning formulation for scientific error detection over academic papers. Following a Reason--Verify--Scan progression, we construct VERA-13K, a 12,900-sample dataset organized into 4,300 matched chains, covering 6 scientific-error categories across the research workflow and broad natural-science domains. We further introduce fine-grained rewards for reasoning completeness, evidence alignment, and error precision. Training Qwen3-VL-8B with VERA-RL substantially improves verifiable reasoning, approaching flagship MLLMs such as Gemini 3 Pro and Qwen3-VL-235B-A22B on Scan.

Subseasonal-to-seasonal (S2S) precipitation forecasting has substantial financial and societal impact, yet remains challenging because of weak predictive signals, high associated uncertainty, and the computational cost of operational systems, which constrains simulation fidelity. We introduce SimCast-S2S, a generative latent-diffusion framework for probabilistic S2S precipitation forecasting that addresses three major bottlenecks in data-driven prediction. First, because S2S prediction requires uncertainty quantification rather than only deterministic point forecasts, SimCast-S2S is the first data-driven system that uses a diffusion-based generative pipeline for S2S prediction, enabling effective sampling from the underlying conditional distribution. Second, since generating large probabilistic ensembles is computationally costly in physical space, SimCast-S2S instead operates in a compact latent space learned by variational autoencoders, enabling efficient large-ensemble generation. Third, diffusion models typically require large training datasets; SimCast-S2S overcomes this via transfer learning with low-rank adaptation (LoRA), pretraining on large ensembles of climate simulations before fine-tuning on limited reanalysis data. On reanalysis data, SimCast-S2S outperforms deep learning baselines, including convolutional neural networks and U-Net architectures. Notably, despite using only a subset of atmospheric input variables and no post-processing, bias correction, or calibration, SimCast-S2S remains competitive with, and in many cases outperforms, state-of-the-art operational systems such as the ECMWF-S2S baseline. These results indicate that latent generative modeling combined with simulation-to-reanalysis transfer learning offers an efficient and scalable path toward data-driven probabilistic S2S precipitation forecasting.

Following clinical decision pathways (CDPs) defined by clinical practice guidelines is essential for safe and reliable medical decision-making. However, existing medical large language model (LLM) benchmarks mainly evaluate final-answer accuracy, providing limited evaluation of models' ability to adhere to guidelines. To address this gap, we introduce MEGA-CDP, a benchmark for evaluating whether medical LLMs can generate guideline-adherent CDPs using provided guidelines as references. MEGA-CDP is constructed from 2,274 English and Chinese clinical practice guidelines through a guideline-to-case pipeline, yielding 42,353 clinical cases with explicit reference CDPs. It supports both single-turn vignette and multi-turn interactive settings, and introduces a CDP-oriented evaluation framework for measuring pathway consistency. Experiments on 16 representative LLMs show that reliable clinical decision support remains challenging for current models, demonstrating the need for CDP-oriented evaluation and the value of MEGA-CDP for advancing guideline adherence in medical LLMs.

Large Language Models (LLMs) generate register-transfer-level (RTL) code with rapidly improving functional correctness. Security of LLM-generated code, however, has been studied mainly for software, where flaws can still be patched after deployment. Insecure RTL offers no such remedy once taped out into silicon. We construct SECRTL-GEN, a multi-language resource-access security benchmark grounded in real SoC IP: 392 tasks over five CWE families and four HDLs (Verilog, SystemVerilog, VHDL, and Python), each with black-box functional and security testbenches. Functional specifications intentionally omit security obligations, matching how obligations are often kept out of functional docs in practice. An empirical study of five frontier LLMs shows a sharp gap: under vanilla prompts they pass functional tests in about 73-79% of cases but security tests in only 14-35%, and stronger functional models are not safer. Adding CWE knowledge raises security, while unaided self-thinking helps less and both security-oriented prompts cut functional pass rates, showing that the bottleneck is missing weakness awareness in the specification, not an inability to write defensive RTL. We present RTL-Obliger, a neuro-symbolic framework that infers these implicit obligations. An LLM extracts a functional-semantic graph from the specification; a symbolic engine then matches it against a CWE pattern ontology to surface mitigation-evidence gaps and signal-level obligations; the LLM finally revises RTL under those obligations in a functionality-preserving two-stage generation. Across five models and four languages, RTL-Obliger raises mean all-pass from 49.6-51.4% (SecV/RESCUE) to 61.6%, with higher security and functional rates than these secure-generation baselines.

Biomedical knowledge graphs (KGs) offer structured medical knowledge that can ground large language model (LLM) reasoning in clinical diagnosis application, yet how KG signal should be integrated into LLMs remains an open question. We present a systematic study spanning five KG task formulations, three training paradigms, two KGs, and three base LLMs. At the task level, all paradigms improve over the non-finetuned baseline, but methods with comparable in-domain accuracy show substantially different knowledge transfer behavior. We introduce Gradient Intervention Density (GID) and Gradient Distortion (GD) to measure how broadly an optimizer modifies the pretrained model. GID and GD together reveal a clear divide: KG-judgment training under KL regularization produces sparse, localized updates (a regime we term as surgical alignment), while task-specific SFT produces dense ones. A controlled ablation shows that the objective and KL contribute to sparsity independently, and the paradigms that produce sparse updates also improve reasoning quality, even when their in-domain accuracy is lower than task-specific SFT. Assessing KG-LLM integration thus requires complementing accuracy with optimization-geometry diagnostics. Our implementation can be found at https://github.com/LARK-NLP-Lab/Surgical-Alignment.

Discrete diffusion models have become a strong, widely adopted class of generators for sequence data, and steering them toward a downstream reward at inference time, without any retraining, is increasingly important. Such training-free steering is done by gradient guidance, by search, or by combining the two. We study the combined regime and identify two weaknesses in how it is usually run: the guided proposal estimates its gradient from a single noisy sample, and the search then resamples particles at a fixed temperature that ignores how rewards spread across each denoising step. We address both with a small set of changes that add no denoiser cost. For the proposal, we lower the estimator variance with a Rao-Blackwellized reveal for differentiable rewards and a leave-one-out baseline for non-differentiable ones; for the search, we standardize the per-step values into a group-relative advantage and prove it collapses to a single active ingredient, an adaptive resampling temperature. We call the resulting method Guided Reduced-variance proposals and Adaptive Selection (GRAS). GRAS is simple yet effective: across regulatory DNA and protein design it attains the best training-free reward, outperforming prior training-free methods and matching or surpassing a reward-fine-tuned model, and it remains effective even for non-differentiable rewards.

Self-evolving language models have recently emerged as a promising path toward superintelligence, with the advantage of reducing the cost of human supervision. While considerable progress has been made in verifiable domains, self-evolution in unverifiable domains remains substantially less explored. We propose Judge co-adaptation from Zero data (J-Zero), a unified Challenger--Solver--Judge co-evolution framework that supports self-improvement across both domains. The Challenger and Solver co-evolve through an adversarial interaction: the Challenger generates increasingly difficult tasks, while the Solver learns to produce higher-quality responses to them. In parallel, the Judge co-adapts using preference pairs whose ordering is known in advance from how each response was produced, i.e., the Solver's answer over the Challenger's, and its decomposed-and-recombined answer over its one-shot answer, rather than from the Judge's own scores. J-Zero outperforms the baselines by an average of 4.2 points on verifiable and 8.0 points on unverifiable domains, and continues to improve through at least ten iterations, whereas the baselines degrade after two.

Low-bit quantization offers a promising avenue for reducing the computational and memory demands of Multimodal Large Language Models (MLLMs). Recent hardware support for low-precision formats, ranging from MXFP8 to ultra-low-bit formats such as MXFP4 and HiF4, has accelerated research into efficient MLLM training and deployment. In this work, we present a systematic study of these quantization schemes in representative MLLMs that span both video generation and reasoning tasks. Our analysis shows that MXFP8 achieves near-lossless performance, whereas aggressive 4-bit quantization leads to significant degradation. Through extensive ablations, we identify activation quantization as the primary source of this performance loss, contributing substantially more than weight quantization. Motivated by this observation, we propose Residual Fallback Quantization (RFQ), a lightweight activation reconstruction framework that supplements the primary ulta-low-bit activation representation with an auxiliary quantized residual pathway. By explicitly modeling and compensating for quantization errors, RFQ improves activation fidelity while preserving the efficiency advantages of ultra-low-bit computation. RFQ requires no architectural modifications and incurs negligible computational overhead. Extensive experiments on Wan2.2 and Qwen3-VL demonstrate that RFQ consistently recovers a substantial portion of the performance lost under the quantization of MXFP4 and HiF4, significantly narrowing the gap to BF16 baselines across both generation and 4 reasoning benchmarks. Our findings establish activation quantization as the dominant bottleneck in ultra-low-bit MLLMs and highlight residual-based activation reconstruction as an effective and practical strategy for robust 4-bit deployment.

Diffusion multimodal large language models (dMLLMs) have recently emerged as a new decoding paradigm for multimodal generation. Starting from a fully masked sequence, dMLLMs progressively decode the sequence by unmasking a subset of the remaining masked positions at each step. Since the selected tokens serve as the prediction context for subsequent steps, deciding which tokens to decode is crucial to the quality of the final output. The most common strategy prioritizes tokens based on a certainty measure that tends to favor tokens frequently observed in the training data. Recent approaches instead order tokens according to their influence on subsequent predictions, but do not explicitly account for the input image. We propose the Visual Information-Guided Sampler (VIG-Sampler), which prioritizes tokens based on their attention to image tokens. We further impose a constraint that penalizes candidate tokens whose image-attention distributions are similar to those of previously selected tokens, thereby increasing the information gain of the decoded subset. Extensive experiments on 7 captioning and VQA benchmarks with 3 open-source dMLLMs demonstrate the effectiveness of VIG-Sampler, which outperforms the Info-Gain Sampler by an average of 19.3 CIDEr points across the captioning benchmarks and surpasses it on COCO Caption while using only half as many decoding steps.

When researchers compare multilingual models for probing, interpretability, or cross-lingual transfer, they often align embedding spaces and assume that shared-language representations are comparable. We show that this assumption can be premature for decoder-only models. We pretrain paired 310M-parameter models (one English-only, one bilingual) across eight typologically diverse languages, separately controlling for English exposure, total compute, and document overlap. After aligning on shared English vocabulary, we test held-out words and find that token embeddings look similar after alignment, but the deeper hidden states that the model uses for prediction do not. This gap holds for all eight languages and survives controls for document overlap and alternative alignment methods. This hidden-state mismatch grows through middle transformer layers, suggesting that it arises from contextual processing rather than the input representations where alignment is performed. Embedding alignment can mask real differences in how models internally represent a shared language, which matters for any downstream study that treats aligned models as interchangeable.

Diffusion large language models (dLLMs) offer a promising alternative to autoregressive generation by decoding multiple tokens in parallel through iterative denoising. However, increasing decoding parallelism often degrades generation quality, as early errors can contaminate later contexts. Revocable decoding mitigates this issue by re-evaluating decoded tokens and remasking unreliable ones, but existing methods overlook that unreliable tokens may also corrupt the verification context itself. We identify this failure mode and propose Dependency-Aware Revocable Decoding (DARD), a training-free framework that separates tokens into masked, candidate, and unmasked states. DARD verifies candidate tokens using a selective context that excludes less reliable tokens and adaptively regulates their influence on subsequent decoding. Experiments across 12 textual and multimodal benchmarks on 3 open-source dLLMs show that DARD consistently improves the speed-quality Pareto frontier over recent revocable decoding methods, achieving a 2.71$\times$ speedup and a 4.35-point CIDEr score gain over Saber on Flickr30K.

Contrastive reinforcement learning (CRL) scales effectively in goal-conditioned tasks by casting policy learning into a self-supervised contrastive objective. However, in a failure-terminated Markov decision process, established CRL considers pre-failure future goals only when constructing positive samples, without accounting for the probability mass removed by failure termination. Our theoretical analysis shows that this omission induces a systematic overestimation bias in goal-reaching values. Consequently, near-failure trajectories provide disproportionately strong supervision of success despite retaining little future occupancy. Unsafe actions can thereby be reinforced through catastrophic failure bootstrapping, leading to failed policy learning and unsustainable goal-reaching behaviours. To address this problem, we introduce two minimal yet strong corrections: mass-weighted InfoNCE corrects the overweighting of short surviving futures in critic learning, and a log-survival-mass score restores the missing survival mass in policy optimization. The resulting method, Safe Contrastive Reinforcement Learning (Safe-CRL), requires only the one-bit signal provided by failure termination to scale safe goal-conditioned policy learning. Across twelve failure-prone robot navigation and locomotion tasks, Safe-CRL consistently improves survival and substantially outperforms the Scaling-CRL baseline in goal-reaching performance. Additionally, deep Safe-CRL policies exhibit complex failure-avoidance behaviours. This study completes the CRL theory under failure termination and provides a scalable safe RL framework. The code is available via https://github.com/RomainLITUD/safe-crl.

Agentic (tool-using) language models are mainly trained on tool-call traces and agent trajectories during post-training. These data provide direct behavioral supervision, but producing them requires task environments, execution, and verification, making broad tool and task coverage expensive. Publicly available skills offer another source of training data: they encode reusable tool semantics and workflows but are typically used only as inference-time context. We introduce Skill Pre-Training (SPT), a mid-training method that applies causal language modeling to SkillCorpus, a collection of public multi-file skill packages, optionally mixed with general data. To preserve relations among files within each package, we also introduce Reference Insert, a reference-aware assembly strategy that places supporting files near their mentions in the primary instruction. Experiments across multiple model scales and post-training recipes show that SPT consistently improves agentic performance over mid-training on general or trajectory data, while largely preserving general performance. Data mixture experiments show additional benefits from combining skill data with general annealing corpora. These results indicate that skill packages are a valuable data source for pre-training agentic language models.

Recent advances in agentic large language models (LLMs) have enabled increasingly autonomous software engineering workflows, yet automatic machine learning (ML) paper-to-code reproduction remains a challenging long-horizon problem. Unlike conventional code generation, this task requires constructing and maintaining a fully functional repository whose state continuously evolves during execution. Existing systems typically rely on static upfront planning followed by sequential file-level generation, which often leads to inconsistencies as dependencies, interfaces, and execution feedback change over time. We propose DeepRepro, a state-aware framework for paper-to-code reproduction based on execution-state-aware subplanning. DeepRepro dynamically transforms evolving repository states and runtime feedback into fine-grained implementation subplans, keeping planning aligned with execution throughout repository construction. The framework further incorporates repository-aware orchestration and a lightweight process-aware interface for transparent monitoring of long-horizon reproduction. Experiments on PaperBench Code-Dev show that DeepRepro consistently outperforms strong scientific and commercial code-agent baselines.

A looped transformer performs inference by iterating a weight-tied map, making its computation a dynamical process whose cost is set by the resulting inference dynamics. Here we show that networks with identical architecture and objective, trained to identical accuracy, nevertheless realize distinct dynamical phases depending strongly on initialization, and that the bifurcation defining each phase determines how test-time compute scales. The phases are distinguished by their bifurcation mechanisms, including a saddle-node fold and a Neimark-Sacker-type transition to bounded nonstationary motion. In the fold phase, a one-dimensional normal-form reduction predicts both the relaxation-time and spectral-gap amplitudes from local derivatives of the trained map, yielding the parameter-free relation $τ(\varepsilon)[1-λ_{\max}(-\varepsilon)]\toπ$. Composed with a regular distribution of problem difficulty, the same critical slowing down produces the workload-level tail $P(τ>N)\sim N^{-2}$. In the Neimark--Sacker phase, the fold scaling law disappears rather than merely changing its prefactor. Thus, test-time compute is not determined by architecture alone. It is governed by the dynamical phase of the solution found by training.

Randomized sketch-and-solve algorithms accelerate overconstrained $\ell_2$ regression by replacing the input with a smaller problem. Standard subspace embeddings guarantee that the cost of the regression is nearly preserved, but coordinate-wise accuracy of the solution is more delicate: we want the solution vector itself to be close to the optimal solution in $\ell_\infty$ norm. In particular, we want to find a vector $x'\in \mathbb{R}^d$ such that $\|x'-x^*\|_\infty\leq \fracε{\sqrt d}\cdot \|Ax^\star-b\|_2\cdot \|A^\dagger\|_{\rm op}$. Price, Song and Woodruff initiated the study of this problem and showed that the subsampled randomized Hadamard transform (SRHT) with $O(ε^{-2} d^{1+Θ(\sqrt{\log\log n/\log d})})$ rows achieves this guarantee. A subsequent work of Song, Ye, Yin and Zhang claimed to improve the row count to $O(ε^{-2}d\log^3 n)$. Unfortunately, their proof relies on an independence assumption that does not hold in general, and we exhibit an explicit instance on which it fails. To achieve a truly nearly-linear-in-$d$ row count, we introduce a new fast, dense randomized transform, which combines a randomized Hadamard flattening, a random permutation, and balanced, disjoint Gaussian pooling. Conditioned on the Hadamard-and-permutation stage, the sketched problem becomes an exact Gaussian regression in which the noise is independent of the entire sketched design; this conditional independence is exactly what the earlier argument was missing. Our sketch yields the $\ell_\infty$ guarantee with $m=O(ε^{-2}d\log d)$ rows, uses one Hadamard pass with a padded internal dimension $N=\widetilde{O}(n+ε^{-2}d^3)$, and is efficient to apply: the sketched pair $(SA, Sb)$ can be computed in $O(Nd\log N)=\widetilde{O}(nd+ε^{-2}d^4)$ time.

Reinforcement learning-based knowledge distillation has the potential to transfer complex reasoning from teacher to student models, yet it currently faces a critical dilemma: researchers must choose between sparse outcome-based rewards, which provide insufficient logical guidance, or expensive neural Process Reward Models (PRMs) for dense signals. We resolve this by introducing SPEAR (Symbolic Process Evaluation and Alignment Reward), a training-free and plug-and-play process reward method for sequence-level on-policy distillation. SPEAR projects natural-language reasoning traces into domain-adaptive symbolic milestones, providing an efficient proxy for process-level reasoning alignment. By utilizing the longest common subsequence (LCS) to align student explorations with teacher milestones, SPEAR provides a dense, order-aware reward signal that enforces logical consistency without the need for an external neural verifier. Our experiments across math, science, and commonsense reasoning tasks demonstrate that SPEAR effectively bridges the reasoning gap between student and teacher models via sequence-level distillation with efficient dense process rewards. Our code and data are available at: https://github.com/zhuochunli/SPEAR.

While Physics-Informed Neural Networks (PINNs) have emerged as a transformative paradigm for solving complex differential equations, their reliance on backpropagation-based gradient descent and automatic differentiation (AD) imposes significant computational bottlenecks and severe non-convex optimization challenges. To overcome these fundamental limitations, we propose the Physics-Informed Stochastic Configuration Machine (PI-SCM), a novel backpropagation-free framework for both forward and inverse problems in differential equations. The core mathematical contribution lies in the analytical evaluation of local Jacobians for nonlinear differential operators, which facilitates a linearized representation of the physical loss and projects it into a unified, linearized algebraic subspace. This reformulation allows for the explicit determination of optimal network weights via a sequence of generalized linear least squares solvers, effectively bypassing the iterative traps of traditional nonlinear optimizers. We develop a progressive algorithmic suite comprising localized construction (PI-SC-I), sliding-window updating (PI-SC-II), and global updating (PI-SC-III), and rigorously establish their universal approximation properties. Extensive experiments demonstrate that PI-SCM achieves high-fidelity predictive accuracy and robust parameter identification while accelerating the training process by orders of magnitude compared to standard PINNs. Our work provides a highly efficient and scalable foundation for next-generation, real-time Scientific Machine Learning applications.

Autonomous agents are increasingly adopted to complete complex, multi-tool workflows in real-world settings. However, existing benchmarks typically separate tasks by application or capability and evaluate agents in environments that are cleaner and more stable than those encountered in practice. We introduce DuMateBench, a real-session benchmark reconstructed from anonymized and privacy-screened user sessions collected from a large-scale production agent platform. Each task preserves the relevant pre-solution interaction history, persistent configurations, and workspace state, and is then validated through human verification. The resulting benchmark comprises 200 tasks spanning 8 broad scenarios and 17 fine-grained capability categories, with most tasks requiring multiple capability coordination. We execute these tasks in isolated Docker containers injected with three forms of real-world environmental complexity: Insufficient, Unstable, and Noisy, and assess performance using a hybrid deterministic and LLM-as-Judge evaluation protocol. Experiments across five representative autonomous-agent frameworks paired with four state-of-the-art LLMs reveal substantial gaps in strict task completion. Complementary robustness, efficiency, and diagnostic analyses further show that performance under environmental perturbations is jointly shaped by the capabilities of the LLM and the surrounding agent framework. The code and data are publicly available at https://dumatebench.com/.

We present Chart2SVG, a multimodal large language model that converts static raster charts into structurally organized, semantically enriched SVGs that support programmatic editing. By incorporating chart-specific semantic tokens into a vision-language model, Chart2SVG captures both geometric primitives and their functional roles. To support robust structural recovery, we introduce Beagle+, a dataset of 33K canonicalized and structurally distilled chart samples. Our approach combines specialized training objectives with a rendering-aware post-training phase, producing SVGs that are both visually accurate and structurally consistent. To facilitate higher-level manipulations, we construct a Chart Structure Graph (CSG) that exposes visual dependencies, enabling tasks such as interactive exploration, chart repurposing, and layout reuse. Experiments show that Chart2SVG substantially outperforms baselines in reconstruction fidelity and downstream editing utility, advancing the development of intelligent and interactive visualization tools.

High-throughput drug screening relies on low-cost primary assays to prioritize compounds for more expensive dose-response profiling, where potency is ultimately quantified. Current screening strategies largely focus on identifying compounds that will confirm biological activity on follow-up, implicitly assuming that confirmed activity will also yield a usable potency estimate. However, confirmed biological activity in screening does not necessarily translate into a quantifiable potency, because active compounds can still fail to produce a reportable dose-response estimate. We therefore present a framework for modeling quantifiability, whether follow-up testing will yield a usable potency estimate, as a distinct triage objective from biological activity. Quantifiability was strongly predictable from the preceding low-cost screen, with most predictive information arising from the observed screening features rather than molecular structure. Response-based predictors remained robust on previously unseen chemical scaffolds and generalized across held-out assay-mechanism families, while the probability of successful quantification varied strongly with response amplitude and assay context. These findings establish experimental measurability, distinct from biological activity, as a predictable property of screening outcomes and show that quantifiability-aware triage can improve the allocation of costly dose-response profiling capacity.

Audio-video generation is rapidly moving from prompt-driven synthesis toward multimodal conditioning, where text, images, audio, and video can jointly shape the generated output. This shift changes the nature of safety evaluation: harmful intent may no longer reside in any single input, but instead emerge from how otherwise benign or weakly harmful conditions interact across modalities and time. Existing safety benchmarks, however, remain largely prompt-centric or tied to fixed conditioning interfaces, leaving such compositional risks difficult to study systematically. To bridge this gap, we introduce Multi2AV-Safety, the first safety benchmark, to the best of our knowledge, to cover all 11 non-singleton T/I/A/V conditioning configurations for audio-video generation, comprising 11,024 attack instances. Evaluation on Multi2AV-Safety reveals systematic weaknesses in representative multimodal safety guards across attack mechanisms and harm-evidence structures. Our evaluation reveals two complementary failure modes: harmful semantics can emerge from the combination of individually benign inputs, while explicit harmful cues can become harder to detect when mixed with benign multimodal context. Together, these results identify \emph{compositional risk perception} as a central capability gap in safeguarding multimodal-conditioned audio-video generation: current safety guards fail to reliably integrate safety evidence across modalities and time, even when all conditioning inputs are observable. The dataset will be publicly released in October 2026.

Long-horizon agent runs generate experience that can improve both the current run and future work. Most self-improvement methods process this experience only after execution ends, so they cannot redirect the active run or immediately apply and validate lessons learned from it. We argue that self-improvement should instead be live, using emerging experience both to redirect the active run and to update the persistent harness. Existing agent architectures do not fully support this goal. Single-agent self-correction combines task execution and trajectory assessment within one context, while subagent delegation separates execution but typically cannot redirect an active subagent. We present PILOT, a supervisor-worker harness for live self-improvement through two coupled mechanisms: (1) live steering lets a separate supervisor redirect or abort the active worker during execution; and (2) live self-evolution distils procedures and failure modes revealed during execution into reusable skills and memory. Across two frozen backbones and three benchmarks, PILOT ranks first in five of six configurations. On Terminal-Bench 2.0, PILOT outperforms counterpart harnesses by up to 9.8 percentage points. In the self-improvement setting, PILOT gains 14.6 points with GLM-5.1 and 12.4 points with Kimi-K2.6. Mean output tokens fall by 42.9% and 47.4%, while successful evaluations per million output tokens rise by 110.3% and 134.0%, respectively.

In this paper, we explore multi-expert Conformal Risk Control (CRC) algorithms for pairwise LLM-as-a-Judge evaluation in open-ended dialogue. Our core insight is that multi-expert aggregation offers a complementary remedy to CRC: whereas CRC controls risk at the decision threshold through abstention, aggregation sanitizes the scoring function at its source. Guided by this, we first design two multi-expert CRC methods: Score Averaging and Decision Voting, which aggregate at the score and decision levels, respectively. While both strategies outperform single-expert methods on homogeneous expert panels, on heterogeneous LLM judges they remain risk-valid but recover only limited coverage, because a uniform threshold cannot match the experts' distinct scoring scales. To resolve this issue, we further propose Marginal-Calibrated Conformal Consensus (MC3): it captures distinct per-expert scales via initial threshold ratios, while jointly tuning a unified decision function $C_t(x)$ applied identically in both calibration and test, thereby preserving exchangeability. To evaluate our framework, we construct Panel, a 1,800-pair human pairwise-preference benchmark for open-ended dialogue. It is built on responses generated by four open-weight LLMs over dialogue contexts from three domains (ESConv, MSC, DREAM), with full logit access. In experiments, we find that both Score Averaging and Decision Voting substantially improve accuracy and acceptance rate on homogeneous panels. Notably, MC3 extends these gains to heterogeneous panels by accommodating distinct per-expert scoring scales across all three datasets.

We establish high-probability bounds for mixed input derivatives of wide random neural networks whose activation derivatives satisfy a factorial growth bound. Our main result specializes these estimates to $\tanh$ networks with Xavier initialization. A direct deterministic analysis based on Euclidean operator norms of the weight matrices yields derivative bounds that generally grow exponentially with the depth. We show that this growth can be substantially improved for sufficiently wide Gaussian networks by isolating the term that is linear in the highest-order derivative and controlling the corresponding tangent directions by measurable finite nets. For scalar-output $\tanh$ networks with Gaussian weights and Xavier initialization, we prove that there exist constants $C,C_0,C_1>0$ such that, whenever the common hidden width satisfies $n \geq C\left(L^3n_0^2(1+\log n_0)+L^2\left(1+\log(L/η)\right)\right)$, then, with probability at least $1-η$, the estimate $\left|D^u\mathcal{R}_{Φ^{(L)}}(x)\right| \leq C_0 |u|! (C_1L)^{|u|-1}\prod_{j\in u}β_j(η,n_0)$ holds simultaneously for every non-empty $u\subseteq[n_0]$ and every $x\in[0,1]^{n_0}$. Thus, the first-order derivative bound is independent of the depth, while a square-free mixed derivative of order $|u|$ grows at most polynomially as $L^{|u|-1}$, apart from the coordinate factors. As consequences, we obtain high-probability bounds for the Euclidean Lipschitz constant and for weighted Sobolev norms of the network realization. The latter connect the derivative estimates to quasi-Monte Carlo integration and indicate how such regularity can enter the analysis of QMC-based training.

Scientific computing is undergoing rapid transformation as advances in artificial intelligence, heterogeneous computing, automation, and data-intensive research reshape not only computational tools but also the institutions, workforce models, and collaborative practices that support scientific discovery. This report synthesizes insights from the 2026 Workshop on Next-Generation Ecosystems for Scientific Computing, the second in a three-year series focused on strengthening scientific computing ecosystems through socio-technical co-design. Workshop discussions identified four interdependent strategic themes: software ecosystems for AI-enabled scientific discovery; trust, validation, and traceability; human-AI teaming and paradigm shifts; and workforce, pedagogy, and governance. The report translates these themes into eight priorities for community action spanning shared research infrastructure, trust and traceability, user experience, human-AI teaming, workforce development, cross-sector coordination, stewardship and sustainability, and evaluation of scientific value. Together, these priorities outline directions for building scientific computing ecosystems that remain trustworthy, sustainable, innovative, and resilient as AI assumes a growing role in scientific work.

Two of the most fundamental questions in statistical learning theory are the following: which prediction problems are learnable, and how should they be learned? For the former, elegant answers often take the form of combinatorial dimensions. The latter question, however, has proved considerably more elusive: all known general-purpose multiclass learners rely on intricate orientations of exponentially large one-inclusion structures, and familiar algorithmic principles such as proper learning and regularization remain poorly understood. Motivated by prior work, we ask whether learning reduces to proper learning---possibly over a larger hypothesis class---and whether proper or improper multiclass learning can ultimately be captured by suitable regularizers. Our primary results answer both questions negatively, resolving three open problems from prior work. First, we exhibit a learnable multiclass problem that cannot be embedded in any properly learnable class, meaning learning cannot be reduced to proper learning by enlarging the hypothesis class. Second, we demonstrate that proper learning can require training error and characterize this phenomenon precisely: every properly learnable class admits a proper learner making $o(m)$ errors on samples of size $m$, but every prescribed sublinear scale $a_m=o(m)$ is necessary for some properly learnable problem. Third, regularization is not a general learner: we exhibit a properly learnable class that cannot be learned by any Structural Risk Minimization (SRM) learner, and a learnable class that cannot be learned by any local regularizer. We complement these impossibility results with a positive theory that gives two sufficient conditions for SRM learnability and characterizes SRM representability through integrability of revealed preferences.

We study online prediction for a specific finite-alphabet, exogenously driven source with infinite input memory. Independent Rademacher inputs $(U_t)$ are observed sequentially, and the next binary mark has logit $\sum_{j=1}^{t}θ_jU_{t+1-j}$, where $\abs{θ_j}\leq r_j$ and $\sum_jr_j\leq B$. Regret is expected cumulative excess log loss. Lag $j$ can affect prediction by scale $r_j$ and enters only $n_{T,j}=T-j+1$ prediction rounds, leading to the lag-resolved spectrum $Γ_T(r)=\sum_{j=1}^{T}\log\!\left(1+n_{T,j}r_j^2\right)$. For every summable envelope, a localized Bayesian mixture proves $\cR_T(r)\leq CΓ_T(r)$. For exponential and polynomial envelopes, under the stated finite-sample dimension condition, a Toeplitz-design converse proves $\cR_T(r)\geq cΓ_T(r)$, with constants allowed to depend on the fixed decay parameters and the logit bound. Thus $Γ_T(r)$ is the minimax cumulative-regret scale for this source class in these canonical regimes, giving $Θ(α^{-1}\log^2T)$ for $r_j=Ae^{-αj}$ and $Θ(T^{1/(2s)})$ for $r_j=Aj^{-s}$, $s>1$. The converse is specific to the exogenous lagged model and is not a profile-only theorem for arbitrary stationary infinite-memory sources. Retaining only the most recent $h$ inputs costs order $\sum_{j>h}n_{T,j}θ_j^2$, yet the same worst-case truncation profile can correspond to polynomially different regret. A scaled online Newton predictor attains the spectrum upper bound.

Large language models often exhibit sycophancy, revising their answers to align with users when users push back. Such answer flips, however, can arise from different causes. One possibility is that the model simply aligns with the user's feedback in order to satisfy them. Another is that the feedback genuinely contains useful evidence, prompting the model to update its answer in a rational way. We distinguish them as Unsupported-Yielding and Rational-Updating. Prior work focuses primarily on suppressing Unsupported-Yielding, while overlooking its effect on Rational-Updating. We address this gap with a two-turn evaluation framework that measures the two behaviors separately. Across representative training-time and inference-time interventions, we find that anti-sycophancy methods often encounter a trade-off in which reducing Unsupported-Yielding can sacrifice Rational-Updating, and vice versa, even when the two objectives are optimized jointly. Mechanistic analysis suggests that the two behaviors share an internal substrate: the MLP neurons and attention heads driving them overlap substantially, and their associated steering directions are positively aligned. We further conduct a preliminary orthogonalized steering exploration, which yields modest, backbone-dependent selectivity gains. Overall, our results suggest that anti-sycophancy should be treated not as a simple suppression problem, but as a selectivity problem, where effective interventions should preserve Rational-Updating while reducing Unsupported-Yielding.

Model merging enables combining multiple fine-tuned models without additional training, but its safety implications remain poorly understood. Prior work primarily attributes merging risks to unsafe constituent models, implicitly assuming that merging individually aligned models preserves safety. In contrast, we show that model merging reveals a previously overlooked jailbreak risk rooted in the pretrained foundation model, even when all constituent models are individually safety-aligned. Motivated by this observation, we study a new threat setting where an attacker constructs jailbreak prompts that generalize across merged models sharing the same pretrained backbone, without access to the exact merging coefficients or constituent checkpoints. To exploit this phenomenon, we propose \textbf{Basin-Aware Jailbreak (BAJ)}, which formulates jailbreak generation as a min--max optimization over the merging space to produce transferable adversarial suffixes across merged model families. Experiments across diverse backbones and merging settings show that BAJ achieves consistently high transfer success rates and remains effective under existing defenses.

Optical Character Recognition (OCR) for text recognition using machine vision has significantly improved, particularly when handling heterogeneous textual data. Traditional OCR models struggle with script variations, writing styles, and degraded documents. Advancements in technology are leading to new AI models with improved architecture for handling multiple languages and complex data formats. Despite this progress, a comprehensive evaluation of OCR advancements remains limited. Based on the established preferred reporting items for systematic reviews and meta-analysis (PRISMA) guidelines, this literature review presents an extensive assessment of OCR research to trace the evolution of AI models over the past decade. It explores the transition in AI models, application domains, data types, linguistic coverage, and challenges. Through a detailed analysis of 97 selected studies published during January 2015 - January 2025, key OCR models are identified, and their performance, strengths, and limitations are analyzed. The findings highlight how OCR technologies have evolved to address structured and unstructured text, scene text recognition, and multilingual processing. Unresolved challenges include limited resources for underrepresented languages, high variability in handwritten text, visual similarity among characters, and constraints in real-time OCR applications. To address these issues, several promising approaches are proposed. Key suggestions include self-supervised learning, multimodal AI, automated machine learning (AutoML), AI-assisted postprocessing, tiny machine learning (TinyML), and the creation of joint corpora for script matching. The future recommendations aim to enhance OCR accuracy and tackle the challenges identified for real-time industrial applications. This study will guide future research and establish a foundation for OCR field.

Navigation in unknown environments to find unforeseen objects has become increasingly feasible with capable vision and language foundation models. However, these models also introduce non-negligible inference latency, which becomes an important concern when agents must operate continuously in the real world. Most state-of-the-art methods are still developed in synchronous simulators, where the environment waits for the agent to act and inference time is effectively free. As a result, agents are often designed around the sequential execution of perception, reasoning, and action, with little regard for time constraints. Under real-time execution, where wall-clock time counts towards the task budget, the inefficiencies of these architectures become clear. We show that recent zero-shot object navigation methods suffer consistent performance degradation under such realistic timing conditions. Motivated by this observation, we propose RTNav, a simple but effective architecture that treats inference latency, asynchronous environment stepping, and bounded compute as explicit design considerations. Evaluated on real-time variants of HM3D-v1, HM3D-v2, and HM3D-OVON, RTNav improves the success rate by up to 11% and the Success weighted by Completion Time by up to 5.1 points over prior work.

Decentralized learning systems aim to collaboratively train models across multiple clients without relying on a central coordinator. While decentralization improves scalability, privacy, and robustness, it also exacerbates three fundamental challenges: statistical heterogeneity across clients, fairness in client-level performance, and stringent communication constraints. This raises a natural question: \emph{how fair can decentralized learning be under limited communication?} We address this question by presenting a unified framework for decentralized learning under communication constraints, bringing together graph-based personalization, agnostic fairness, and compressed event-triggered communication. Specifically, we propose a new algorithm DMFL-SQ, a decentralized multi-task learning algorithm that couples personalized model training over a communication graph with an agnostic mixture fairness objective, while reducing communication through sparsification, quantization, and event-triggered synchronization. We establish convergence guarantees for general non-convex objectives and show that DMFL-SQ achieves an $\mathcal{O}(T^{-1/2})$ rate in expected squared Moreau-envelope stationarity despite sparse, quantized, and event-triggered communication. We further derive PAC-Bayes generalization guarantees for the fairness-aware mixture objective. Experiments on CIFAR-10 and the real heterogeneous MUSMET EEG dataset demonstrate that DMFL-SQ substantially reduces communication while maintaining predictive performance and improving fairness across clients. Together, our theoretical and empirical results show that personalization, fairness, and communication efficiency can be jointly achieved in decentralized learning while preserving the dominant convergence rate.

Point-cloud data routinely captured by modern imaging and sensor technologies provide detailed geometric descriptions of objects and environments, but their analysis is hindered by large data volumes, localization noise, and missing information. In addition, existing point-cloud reconstruction pipelines typically return a single best-fit structure without uncertainty quantification. We introduce a fully Bayesian framework for representing point-cloud data and reconstructing closed curves, in which observed points are modeled as noisy perturbations of latent locations constrained to lie on the underlying curve that is regularized by a non-parametric prior. Posterior inference in our framework is carried out using a series of Markov chain Monte Carlo samplers tailored to point-cloud characteristics. Numerical experiments, including synthetic examples and real-world LiDAR datasets, show accurate reconstructions and quantified uncertainty over the recovered curves.

When a single policy is trained in parallel across multiple environments of the same task, such as procedurally generated levels, randomized dynamics, or curricula, implementations commonly use one critic across all sampled environments. Yet different environments can assign different expected returns to the same input visible to the critic. A critic without environment information must then reconcile distinct value targets, systematically shifting the sampled advantages within individual environments. Using illustrative bandit models with multiple environments and a common optimal arm, we characterize how this value mismatch redistributes sampled policy updates, reinforcing unhelpful actions while attenuating or even reversing useful ones. The oracle processes using no baseline, the shared value, or the value specific to the sampled environment have the same mean logit update at a fixed policy and converge to the same optimal policy, yet their realized learning paths can differ sharply. The analysis motivates a minimal intervention: give only a logged environment index to the critic so that it can separate the value targets. Controlled CartPole and MuJoCo experiments expose the predicted shifted values, advantages, and performance gaps. In the more complex BipedalWalker and Procgen settings, the same intervention yields more stable learning and higher returns. Across all $16$ Procgen games, the multihead conditional critic improves aggregate normalized return on $600$ unseen levels per game by $40.8\%$. In conclusion, the theory identifies value mismatch as a direct mechanism through which critic sharing can degrade stochastic learning dynamics, not captured by scalar estimator variance alone, and the experiments show that conditioning on an index is broadly effective in parallel reinforcement learning.

Multi-agent large language model systems are widely reported to beat single-model baselines, but the evidence is mixed, and comparisons are usually confounded: pipelines change token budgets, tool calls, and prompts simultaneously, so an aggregate gain rarely reveals what actually helped. We investigate the effect of introducing the manager-worker scaffold over a shared filesystem workspace, with no training and no per-benchmark tuning, measured against the same model answering in a single pass. Across nine models -- five open-weight, spanning 9B to ~2.8T parameters, and four frontier closed models -- on the 100 latest hard LiveCodeBench problems, the scaffold's benefit is real but conditional: large and statistically significant for some (Qwen3.8-27B +23.4, GPT-5.6-Luna +10.6 and GPT-5.6-Terra +8.0, each over five paired passes; Kimi-K3 +30.4 and Minimax-M3 +11.0 over five paired passes with reasoning off, both at $p < 10^{-4}$, and +42 and +12 in a single pass at a 128k cap) and null or negative for others (Qwen3.6-35B -1 to -9 with reasoning off). With the manager, Opus-5 achieves the highest score in the study at 91% in one pass. Running a manager roughly triples the token bill, but it buys accuracy more cheaply than moving to a larger model does: GPT-5.6-Terra with a manager nearly matches Fable 5's single-call accuracy (85.0 against 87.4, $p = 0.59$) at a fifth of the price (\$11.71 against \$61.11 per 100-problem pass, $p < 10^{-4}$), and the Qwen-27B arm does it for \$51.75 on weights anyone can self-host. Our transcript analysis finds several mechanisms behind the gains, of which two recur: context management, in which short worker calls and shared notes organize state and reduce truncation, and problem decomposition. Improvements are modest for large models with reasoning enabled, but larger for some models with reasoning disabled and for smaller models with reasoning enabled.

In many reinforcement learning (RL) domains, environments are connected by prerequisite relations, such as difficulty-increasing edits or parameter increments, which induce a directed acyclic curriculum graph (DAG). Although this structure is often exploited only implicitly, explicitly modeling it can improve training. We introduce PATH, a curriculum-learning framework that performs active learning over the curriculum graph. PATH first expands coverage by sampling diverse curriculum paths and then reallocates training toward regions that remain unmastered. Experiments across diverse environments show that PATH explicitly leverages the graph structure to achieve strong robustness and generalization.

Compositional generalization is usually evaluated through model accuracy. We instead ask which structural or lexical identifications make held-out COGS examples admissible from the structures observed in training. Sentences are represented as functors from syntactic addresses to lexical tokens, and selective collapses induce Kan extensions that propagate observed associations. Across 21 COGS generalization types, admissibility follows distinct identification profiles, while residual failures separate unsupported structural templates. These data-side diagnoses characterize what the training corpus licenses under specified identifications, without training a predictive model.

Finetuning has become the gold standard for refining existing behaviors and inducing new ones in language models, yet it often remains unclear exactly which behaviors emerge during this process. As models grow ever more capable, understanding finetuning better becomes increasingly important, particularly since unwanted behaviors may arise during finetuning. In this paper, we introduce Diff Mining, a simple yet effective framework for identifying what a finetuned model has learned by comparing its logits to those of its base model. Diff Mining effectively surfaces salient tokens that are amplified in the finetuned model, serving as a fingerprint of its training -- even on text unrelated to the finetuning domain. Unlike many existing model diffing methods which require model internals, Diff Mining only needs access to output logits and scales to large models. The framework consists of two modular stages: (i) extracting per-context logit differences between the finetuned and base models on a reference corpus, and (ii) aggregating the resulting signals to construct an interpretable token set representing the finetune. For aggregation, we explore both a simple Top-K frequency method and a Non-negative Matrix Factorization (NMF)-based approach for disentangling multiple finetuning objectives into distinct token clusters. Empirically, Diff Mining succeeds across diverse settings: on finetune domain detection, it significantly outperforms state-of-the-art model diffing methods both in identifying relevant tokens and in downstream performance when an interpretability agent is given access to the extracted token set; on models with injected biases, it identifies more than one third of the biases without targeted probing. Overall, our framework shows promise in developing auditing tools to detect finetuning objectives.

Distributed training across a wide area network (WAN) is challenging, as continuous parameter exchange by islands of compute is constrained by limited bandwidth, high latency, and uneven topology. We propose making the network an active participant in training. On the systems side, such networks should leverage (i) multicast technology to replicate outbound traffic and (ii) in-line FPGAs to aggregate inbound traffic, to ease egress and ingress bottlenecks. These technologies are used for training across workers within a data center, but this paper extends them to the WAN. On the algorithms side, we develop an optimization framework that produces rich synchronization schedules (namely, rotating cliques of islands) around the underlying network topology and these technologies, to maximize information exchange. Finally, we illustrate this on a nine-city topology modeled on the DoubleZero network, a live programmable WAN equipped with both technologies, and show how the optimal schedules shift with the network's capabilities. Together, these can narrow the gap to the gold standard of colocated training.

Bike-sharing systems are an important component of low-carbon urban mobility, but continued expansion creates challenges in both cold-start prediction and equitable resource allocation. Newly deployed stations lack historical ridership records, causing a mismatch between training and inference for graph-based models on evolving networks. Historical demand may also encode structural inequalities, as lower ridership in low-income neighborhoods can reflect limited infrastructure access rather than weak latent demand. Models trained directly on such data may therefore reinforce existing mobility disparities. We propose FairGIN, a fairness-aware graph neural network for demand prediction in expanding bike-sharing systems. FairGIN integrates three components. Expansion-Simulated Increment Training stochastically simulates network expansion during training to reduce the cold-start distribution gap. Attention-Based Knowledge Transfer combines station-adaptive temperature scaling with orthogonal embedding alignment to transfer representations from data-rich existing stations to data-sparse new stations. Fairness-Aware Optimization introduces income-stratified regularization and an equity-calibrated deployment score to support more inclusive station placement. Experiments on NYC and Seattle demonstrate that FairGIN achieves state-of-the-art predictive accuracy across diverse expansion scenarios while substantially reducing income-based disparities without compromising overall system efficiency.

Byte-level BPE tokenizers that use the HuggingFace ByteLevel pre-tokenizer inherit GPT-2's word regex, where a word is defined as \p{L}+, one or more Unicode letters. In abugida scripts, vowels are written as combining marks; this pattern therefore splits each word at every vowel sign. Since BPE merges only within a pre-token, those splits persist through training regardless of vocabulary size or corpus composition. We formalise this effect as a training-free lower bound on fertility. Across 26 languages from a parallel corpus, every one of the 17 abugidas is affected, ranging from 1.47x (Tibetan) to 9.02x (Thai), whereas Latin, Cyrillic, Hangul, and Han show exactly 1.00x. For 5 languages, matched tokenizer pairs that differ only in this character class fall within 2.2% of the predicted floor, scoring 4.78 versus 1.58 tokens per word on Nepali. When the Nepali share of the training corpus is swept from 5% to 95%, the broken tokenizer barely shifts at all (1.7%) while the fixed one shifts 33.9%, which separates a structural ceiling from a data shortage without needing to inspect any code. We train three 268M models that differ only in their tokenizer; the fixed variant achieves 4.43% lower held-out Nepali bits per byte at equal compute, and it still leads when given the same bytes with 1.59x the compute. A census of 3,479 HuggingFace repositories finds the letters-only word class present in 63.3% of the most-downloaded text-generation models, accounting for 72.5% of their downloads. GPT-4o's o200k pattern already uses a mark-aware word class, making the repair itself prior art. We quantify its value, show how to recognise its absence from symptoms alone, map which scripts it reaches, measure how widely it is deployed, and release a 65,536-entry Nepali-English tokenizer with a harness that regenerates every number here from public data on a laptop.

Recent advances in Large Language Models (LLMs) have shown that increased inference-time reasoning can improve performance on complex tasks. However, many existing approaches rely on fixed or preallocated reasoning controls, such as fixed token budgets, pre-execution difficulty estimates, or activation-space interventions, and are often evaluated on standalone reasoning benchmarks rather than full agentic workflows. These assumptions may not hold in agentic AI systems, where reasoning requirements evolve dynamically through planning, tool use, memory retrieval, and agent-to-agent interactions. Consequently, reasoning can become either excessive or insufficient, resulting in unnecessary computation, increased latency, planning drift, excessive tool use, or incomplete solutions. We argue that a major challenge for next-generation agentic AI is not merely how much reasoning a language model should perform, but how it should allocate reasoning according to evolving task demands. We characterize over-reasoning and under-reasoning as recurring failure modes of misallocated reasoning and evaluate them on MATH-500 and the GAIA public validation benchmark. Using tool-decision latency, token consumption, token-limit exhaustion, and answer correctness, our results suggest that cases classified as over-reasoning are associated with higher computational cost without proportional accuracy gains, whereas cases classified as under-reasoning are consistently associated with incorrect or incomplete solutions. These findings motivate future research on adaptive reasoning mechanisms for agentic AI.

Graph neural networks (GNNs) can exhibit unfair behavior even when sensitive attributes are excluded from node features, because graph topology and message passing propagate group-correlated signals under sensitive homophily. Existing fairness-aware GNN methods mainly constrain representations or prediction distributions at a global level, without explicitly controlling the local structural pathways through which biased information propagates during aggregation. We propose Subgraph Filtering for Fair Graph Neural Networks (SF-GNN), a lightweight and architecture-agnostic framework that mitigates structural bias at its source. SF-GNN identifies bias-prone edges by combining sensitive homophily with structural propagation amplifiers, including hub participation and triadic closure. It then incorporates stochastic edge filtering into each message-passing step to selectively downweight or remove these edges while preserving the remaining graph structure. Training further incorporates a statistical-parity regularizer with a warm-up schedule to stabilize optimization. Experiments on five benchmark datasets show that SF-GNN achieves consistent fairness improvements while maintaining competitive predictive performance, leading to a better fairness--accuracy trade-off than recent fairness-aware GNN baselines.

Neonatal mortality risk prediction from bedside monitoring data remains challenging due to extreme class imbalance, heterogeneous clinical risk factors, multi-scale temporal dynamics, and substantial missingness. We propose NeoTriFuse, a reliability-aware multimodal fusion framework for missingness-heterogeneous neonatal monitoring data. Unlike conventional multimodal approaches that treat missingness primarily as a preprocessing issue, NeoTriFuse models missingness as an explicit reliability signal that dynamically modulates modality contributions during fusion. The framework integrates static perinatal variables, local-global temporal encoders, and patient-level statistical summaries through reliability-guided gating mechanisms, while jointly optimizing mortality prediction and an auxiliary length-of-stay objective. NeoTriFuse achieves competitive performance, with an F1 score of 0.6736 +/- 0.0216 and an AUROC of 0.9454 +/- 0.0056. Ablation studies indicate that the local-global temporal architecture and patient-level summary branch contribute most substantially to predictive performance, while reliability-aware gating provides additional improvements on threshold-dependent metrics under heterogeneous observation completeness. Sensitivity analyses further suggest stable performance across nearby hyperparameter settings. Overall, the findings support reliability-aware multimodal fusion as a practical approach for neonatal mortality prediction under realistic clinical missingness conditions.

Code-switching is pervasive in bilingual African conversation, yet most ASR systems assume monolingual input and are evaluated on curated monolingual benchmarks. We present AfriSwitch, a 61.36-hour human-transcribed benchmark of in-the-wild code-switched speech spanning 16 African languages and language varieties, released with switch-level English span tags, perutterance Code-Mixing Index (CMI), and switch-point counts. Corpus statistics show that mixing behaviour varies widely across African languages along two largely independent axes: how often speakers alternate, and how balanced the mixture is. No single scalar captures how code-switched a language is. Benchmarking five open and commercial multilingual ASR systems zero-shot yields word error rates far above published monolingual figures for the same languages, with the best system averaging 35.93% WER and no system falling below 24% on any language. Africa-targeted training, not model scale or nominal language coverage, best predicts performance.

Data-driven prognostics and health management has emerged as a key enabler for Industry 4.0, yet the development of robust remaining useful life (RUL) estimation models is often limited by the scarcity of run-to-failure data. While federated learning offers a promising paradigm to collaboratively train predictive models without sharing sensor data, research efforts have operated so far in the absence of a common evaluation framework. To address this gap, this paper introduces FedCMAPSS, a benchmark for federated RUL estimation based on the commonly-used NASA C-MAPSS dataset. We define a set of five standardized tasks designed to simulate real-world industrial challenges, ranging from ideal IID settings to extreme statistical heterogeneity, and conduct a systematic evaluation of state-of-the-art federated optimization algorithms across multiple neural architectures. By establishing reproducible baselines and making the source code and data splits publicly available, this work aims to provide a standard foundation for developing and comparing federated predictive maintenance solutions.

Voice agents must call tools and hold multi-turn dialogue entirely through speech, yet the dominant paradigm trains them in text. Existing frameworks either cascade TTS and ASR around a proprietary voice API, where gradients cannot flow and per-call cost makes on-policy reinforcement learning prohibitive, or stay in text: they measure voice agents but cannot improve them. We present SpeechGym, an audio-native agentic environment in which two omni-modal models converse in native audio, with no external ASR or TTS and no API boundary, over the unmodified tasks, tools and success check of an established text agentic benchmark, so that the interaction modality is the only variable and the loop stays local and trainable end to end. Audio agentic capability does not follow from audio understanding. The failures speech introduces are perceptual rather than reasoning deficits: the agent picks the right tool and the right argument slot but fills it with a value misheard from the waveform, and that single error cascades into a failed call, a retry of the same call, and a wasted step budget. A second failure is behavioural: under an insistent caller the agent performs an unauthorised write and ends the episode believing it helped. Both are trainable, because the environment labels them for free: a call with a misheard argument fails against the database while a correct one succeeds. The obstacle is sparsity, not signal. Outcome-only GRPO is gradient-starved here, since almost every rollout group fails identically, while a per-turn process reward crediting each successful tool call restores variance to nearly every group. Trained this way, the agent transfers with no further tuning to an independently implemented voice benchmark, more than doubling task success and carrying an open-weights model from last place to second on that leaderboard, while using fewer turns and tokens than before training.

Parameter-Efficient Fine-Tuning (PEFT) strategies such as Low-Rank Adaptation (LoRA) are effective solutions for fine-tuning large-scale pre-trained models; however, their memory requirements scale with the size of the model, $\mathcal{O}(dr)$, where $d$ is the model's hidden dimension and $r$ is the rank. Our proposal, FrameFT, models the parameter update $ΔW$ with a sparse coefficient matrix in a Fusion Frame basis. Fusion Frames can be generated algorithmically and shared across model layers, enabling very efficient updates. Only the sparse coefficients of the basis expansion are stored/optimized, reducing the memory footprint. The sparse structure of the coefficient matrix in FrameFT and the sparsity in the Fusion Frames give large compute benefits, and our analysis provides formal convergence results. We evaluate the idea across a suite of supervised fine-tuning benchmarks, focusing on language tasks, but also report application to vision models. Our experiments show that FrameFT achieves performance on par with/exceeding state-of-the-art PEFT techniques, but needs far fewer trainable parameters.

This paper proposes a framework for constructing a classifier as a safeguard layer, and for developing a complementary diagnostic that identifies which of the classifier's confident decisions can be trusted. This framework, the Latent Diagnostic Taxonomy, consists of (i) constructing a dimensionality-optimized classifier, in which the embedding dimensionality is empirically selected via cross-validated performance rather than fixed a priori, (ii) locating a relatively small set of latent support vectors (~ 29% of total training examples) representing influential prompts for identifying tokens that alter the classifier's predicted labels, and (iii) utilizing such tokens and their associated attack magnitudes for constructing a diagnostic taxonomy. This diagnostic taxonomy provides an end-to-end guideline for flagging prompts that require different treatments: rely Safely on the classifier's decision; flag Heuristic Bias and Heuristic Override cases; route Insufficient Context cases for further human/safety review. Applying the framework to a classifier trained on a public prompt injection dataset, we find that a substantial fraction of its confident decisions (~ 77%) are not robust to removing a single token, and that this brittleness separates into two distinct failure patterns: a confidence calibration failure and a genuinely exploitable shortcut. For each zone of the taxonomy, we also recommend strategies for remediating diagnosed prompts. We illustrate the framework as a series of steps, demonstrating how each step operates.

Protein language models (pLMs) encode information about protein sequences which enable downstream tasks such as structure prediction, but their internal representations are not well understood. Sparse autoencoders (SAEs) provide a promising tool to disentangle latent pLM representations into interpretable features, but existing annotation pipelines largely rely on protein-level annotations derived from database labels and LLM annotations of top activating sequences. Such annotations can overlook the localized residue-level and geometric patterns encoded by sparse features. We introduce an automated and scalable method for interpreting SAE features in ESM-2 by using geometrically inspired features of the protein $\text{C}_α$ backbone. Across ESM-2 8M layers, an FDR-controlled discovery analysis shows that local geometry is significantly associated with many SAE features, with varying levels of predictive strength, expanding coverage beyond database and sequence-based methods. In particular, geometry can distinguish SAE features sharing the same database annotation, revealing substructure within known biological labels. A significant portion of SAE features activate on unannotated metagenomic protein sequences enabling us to use our SAE annotations to better understand these sequences. In addition, ablation experiments at the level of contact prediction show that removing found geometric features shifts ESM-2's predicted contact maps in the direction of the descriptor. This provides a robust method of annotating proteins activated within SAE neurons at a residue level, providing a bridge between mechanistic interpretability and structural biology.

Modern AI workloads and the hardware that runs them evolve on different timescales: architectural definition precedes volume silicon by years, while target workloads shift in months. Design decisions are therefore committed under deep uncertainty and paid for twice, once in the generality added as a hedge, and again when new workloads map poorly onto frozen silicon. As Moore's Law stagnates, specialization is the main remaining source of performance-per-watt and demands a design cycle that runs at the cadence of the workloads. We present an end-to-end AI system that collapses the software-to-silicon stack into a single optimization loop, where hardware and software are co-designed and verified under one objective. Its first demonstration is Redwood, a frontier AI accelerator built for single-batch, low-power, ultra-low-latency inference for physical AI. From a high-level specification by two human architects, the system autonomously generated the performance model, RTL design, UVM environments, formal proofs, firmware, and kernels in under two weeks with no human intervention below the specification. Every block reached 95% coverage via commercial EDA tools, our proprietary formal engine, and hardware-in-the-loop validation. Specification changes were reverified and redeployed to hardware in under 48 hours. Redwood Nano, its ultra-low-power FPGA variant, runs multi-billion-parameter models like Llama and Qwen. Projected onto Samsung 8 nm, the Jetson Orin Nano's process class, Redwood delivers 1.75x the throughput at 1.9x lower power, a 3.4x performance-per-watt gain against a measured Jetson baseline on the same models. Qwen running on Redwood also helped design next-generation Redwood, an early step toward recursive self-improvement. To our knowledge, this is the first production-worthy AI accelerator designed end-to-end by an AI system and running a modern AI model.

Metasurface design increasingly requires fast models that can operate across structurally distinct device families, rather than retraining a separate surrogate for every geometry class. Conventional neural network surrogates often depend on fixed-dimensional descriptors, family-specific output formats, and repeated architecture tuning, which limits their scalability across heterogeneous meta-atoms. Here, we present a unified large language model (LLM) workflow for multi-family metasurface modeling and inverse-design. Geometries, design parameters, and optical response channels were converted into a shared instruction-following text format and used to fine-tune Gemma-2-9B across 8 metasurface families. Compared with single-family baselines, the joint model simultaneously predicted the optical responses of all metasurface families while reducing the MSE for each family by an average of 56.5%. The same representation was also used for inverse design. These results show that a shared sequence-based LLM interface can provide a practical route to cross-family metasurface design while reducing the need for task-specific surrogate architectures.

Medical guidelines encode rich, evidence-based decision logic, yet the specific decision artifact a clinician needs is hard to locate within a guideline, let alone across guidelines covering plausible diseases and treatments. While guideline passages have supported end-to-end question answering, flowcharts remain largely underused in decision support despite their ability to encode actionable clinical pathways. We therefore introduce Case2Flow, a task designed to retrieve the most relevant guideline flowchart for a given patient case from a collection of guideline documents. To support it, we construct FlowAtlas, a curated corpus of 202 flowcharts extracted from 2,080 medical guidelines, together with a pipeline that synthesises 1,911 aligned case-flowchart pairs. Our evaluation of multimodal retrieval methods reveals systematic failure modes, including overreliance on keywords and spurious token-patch matches induced by uninformative background regions in flowcharts. Motivated by this, we propose CRISP, a training-free scoring method that sharpens late-interaction retrieval by suppressing uninformative patches, discounting ambiguous token matches, and incorporating bidirectional query-image alignment. CRISP improves Recall@1 by up to 18.71 percentage points, while a blinded physician assessment on published case narratives provides preliminary feasibility evidence beyond synthetic queries.

Recent work in fair division has focused on either simultaneously satisfying closely related fairness notions or achieving a single notion across the ex-ante and ex-post worlds. We study the compatibility of two fundamentally different fairness notions: envy-freeness and equitability. For indivisible goods-only and chores-only settings, we study the existence and complexity of simultaneously satisfying their relaxations, revealing sharp contrasts between the two settings. We show that EF1+EQ1 may fail to exist even for normalized, additive valuations. Our main algorithmic result computes an EF1+EQ1 allocation for normalized binary goods with at most seven agents. In sharp contrast, binary chores admit the stronger EFX+EQX guarantee for any number of agents, even without normalization. We further initiate the study of cross-notion ex-ante--ex-post guarantees, asking whether randomized allocations can provide ex-ante guarantees for one notion while preserving ex-post guarantees for another.

LLM-assisted decompilation improves readability and re-executability, but still underperforms on stripped C++ functions that use the Standard Template Library (STL). Compilation, optimization, and symbol stripping remove or obscure source-level semantics such as container types and library-call structure, while traditional decompiler output often fails to recover them. We present STILL, a structured semantic interface that predicts function-level STL container semantics from stripped control-flow graphs and renders them as compact hints for LLM refinement. On StlBench, STILL predicts common container-level STL semantics, with the strongest cross-dataset results for stable string and vector slices. On stripped HumanEval decompilation, these hints enable DeepSeek-chat refinement to reach 28.4% executability, compared with 17.4% for no-hint refinement and 8.9% for raw Ghidra decompilation; hint utility is downstream-backbone-dependent, with decompilation-specialized models requiring lightweight adaptation to benefit from the same interface.

Generated computer-vision code can be runnable without satisfying the task contract enforced by a downstream evaluator. We study that gap with Spec2Vision, an experimental framework for producing and evaluating specification-grounded CV pipeline bundles through a staged runtime that keeps the task contract explicit across synthesis, screening, testing, and bounded repair. The benchmark evaluates 17 CV tasks, 10 executable conditions, and 5 repeats per task-condition cell, for 850 primary runs. In the primary 850-run evaluation, Spec2Vision reaches 81/85 evaluator-test passes; removing structural repair drops to 55/85, compatibility scaffolding to 58/85, and generator preflight to 39/85. The executable single-agent baselines expose progressively richer task specifications to the model, culminating in direct source-spec exposure, yet remain much weaker overall, from 17/85 for lightweight task grounding to 35/85 evaluator-test passes. The lightweight baseline nevertheless remains core-runnable in 85/85 runs but reaches only 17/85 evaluator-test passes and 6/85 strict-delivery successes, showing that runnability is not equivalent to delivery. Across this benchmark, the strongest evidence comes from keeping the task contract explicit across staged generation, checking, and repair. Artifacts are provided to support audit of run bundles, model-visible inputs, and derived tables.

Large Language Models (LLMs) are increasingly used to generate executable software systems from natural language specifications, accelerating development and reducing manual implementation effort. Although recent studies have investigated the functional correctness, security, maintainability, and robustness of LLM-generated code, little is known about the long-term reliability of such systems under sustained execution. In this paper, we experimentally investigate software aging symptoms in LLM-generated service-based applications across different programming languages. Using backend scenarios derived from BaxBench, we generated applications targeting JavaScript, Python, and Rust through LLM-based generation platforms, validated them with BaxBench-derived tests, and subjected them to 48-hour workload executions. We monitored memory usage, response time, and throughput and analyzed them using the Mann--Kendall test and Sen's slope estimator. We further complemented the runtime evaluation with static analysis of the generated source code and an exploratory comparison with human-written implementations of related backend scenarios. The results show that memory usage is the most consistent indicator of potential software aging, with statistically significant upward trends in most application-language combinations, while response time and throughput exhibit more heterogeneous behavior. Static analysis identified plausible code-level aging mechanisms, and the comparison with human-written systems showed that aging trends can also emerge in manually developed implementations. These findings indicate that functional correctness alone is insufficient to assess the operational reliability of LLM-generated software before deployment in continuously running environments.

SVD-based low-rank compression has become a fast-growing direction for reducing the memory and computational cost of large language models (LLMs). However, meaningful comparison across existing studies remains difficult as prior evaluations use varied benchmarks, inconsistent ratios, and diverse setups, often failing to isolate low-rank effects from auxiliary techniques. As a result, it remains unclear whether reported gains reflect method-level improvements or differences in evaluation protocol. This lack of comparability highlights the need for a unified, reproducible evaluation platform. To address this problem, we present LowRankArena, a standardized evaluation platform for SVD-based LLM compression. LowRankArena unifies task versions, uniform-precision compression budgets, comparison regimes, and inference measurements, and provides a reproducible pipeline with over 3 TiB released compressed checkpoints. Using LowRankArena, our aligned audit of five representative SVD methods reveals that prior findings are highly conditional under standardized protocols: clear leaders and performance tiers shift across backbones and keep ratios, multiple-choice accuracy can hide large perplexity degradation, and nominal low-rank savings yield workload-dependent and often limited end-to-end speedups. Our code is available at: https://github.com/Zishan-Shao/lowrankarena.git.

Retrieval-augmented methods improve factual accuracy by grounding language models in external knowledge, but retrieving over unstructured text often introduces irrelevant context and offers limited control over the retrieved information. Structured knowledge bases offer a more controllable alternative, yet they are expensive to construct and often brittle to reason over. To address these limitations, we propose KBevo: a co-evolving framework that jointly learns to construct a structured knowledge base and reason over it for knowledge-intensive question answering. By optimizing both components end-to-end with QA outcome rewards, our method enables reasoning success to directly improve the quality of the constructed knowledge base. This leads to larger, better-connected knowledge structures with higher answer reachability, while also improving compositional factual reasoning and controllability compared to standard retrieval baselines.

Volume-based accuracy rewards retrieval-augmented generation (RAG) systems for guessing: a system that answers everything outscores one that declines when its knowledge base cannot support an answer. Building on the confidence-target analysis of Kalai et al. (2025), we present a penalty-aware evaluation framework for deployed RAG products, combining (i) asymmetric scoring (correct +1, wrong -4, abstain 0), (ii) knowledge-gap canaries, questions whose answers are verifiably absent from the knowledge base, so that any answer constitutes ungrounded generation from parametric memory, and (iii) a failure-attribution pipeline that separates retrieval, generation, and abstention-policy failures. Applying the framework to three commercial RAG systems and a no-retrieval baseline on SimpleQA-Verified (1,000 questions x 3 repeats, graded blind by a cross-family three-judge panel with 98.9% unanimity), we find that accuracy when answering is closely clustered across systems (97.0-98.0%), while canary violation rates differ roughly sixfold (16.7% vs. 98.1%). The systems are separated less by what they answer correctly than by whether they answer at all when they should not, and penalty-aware scoring reorders the volume-based ranking accordingly; the reordering is stable across penalty settings from k=1 to k=9. All code, configurations, transcripts, and judge votes are released for independent audit.

Retrieval evaluation for retrieval-augmented generation (RAG) is increasingly designed around whether retrieved passages contain evidence that can support generation, rather than topical relevance alone. We study whether this closer alignment with downstream evidence needs also makes retrieval evaluation more useful for the decisions built from it. Across five retrieval benchmarks and an end-to-end TREC RAG 2025 setting, we examine an answer-support signal in four roles: comparing retrievers, guiding retrieval training and system selection, predicting downstream answer quality, and filtering the evidence supplied to a generator. The signal changes retrieval rankings, but its downstream value is not uniform. It does not reliably improve retriever training; the benefit of using it for system selection depends on how the generator is instructed to use the retrieved evidence; and retrieval scores based on it do not robustly predict answer quality on unseen topics. In a direct evidence intervention, human annotators confirm that filtering preferentially preserves passages containing useful answer evidence, yet different answer evaluators reach different conclusions about whether the resulting answers improve. These results show that making retrieval evaluation more closely reflect the evidence needed for generation does not by itself make every downstream use of that evaluation more reliable. RAG evaluation methods should therefore be assessed with respect to the particular comparisons, decisions, and conclusions they are intended to support.

Standard machine-learning training minimizes a loss function over a dataset, but does not guarantee that the resulting model will satisfy predefined rules or constraints on its outputs. In many real-world applications, ranging from autonomous systems to network routing, such guarantees are essential. We propose CG4AI, a framework that builds a convex combination of AI models while enforcing linear constraints on the combined output. A master linear program (LP) determines the optimal mixture weights, while a pricing subproblem generates new models guided by LP dual variables, focusing attention on the most violated constraints. A cutting-plane procedure extends feasibility guarantees beyond the training set. We apply CG4AI to two problems: (i) digit classification on MNIST, where we demonstrate four distinct uses of constraints, learning from constraints alone, improving adversarial robustness, correcting misclassified examples, and enforcing output relabeling; and (ii) the multi-commodity flow problem, where link capacity constraints are enforced on neural-network routing predictors. Experiments on MNIST and standard SNDLIB benchmark networks show that CG4AI reliably produces feasible predictors while achieving better accuracy than single-model baselines.

Diffusion language models (DLMs) generate text by iteratively denoising masked sequences, but standard decoding either fixes the sequence length or relies on ad hoc stopping rules, often leading to unnecessary denoising steps. We recast length selection as a discrete-time survival problem over the end-of-sequence token and propose a plug-in, training-free length predictor that can be added to any existing DLM. Across reasoning and code-generation benchmarks, survival-guided length decoding speeds up inference by up to 7 times while preserving task accuracy. We further find that predicted lengths vary widely even within the same dataset, making model performance sensitive to the chosen length.

Large language models are increasingly deployed as autonomous agents serving users on behalf of companies, placing them in settings where user and deployer interests can conflict. When an agent knows that a user is owed something its deployer would prefer to deny, does it remain honest? Answering this is difficult because false statements can reflect either ignorance or hallucination rather than deception. To address this challenge, we introduce KnownLieBench , a knowledge-verified benchmark that first confirms through a neutral probe that an agent knows a user's entitlement, and then evaluates whether it makes false claims once an incentive to deny that entitlement is introduced. Specifically, KnownLieBench covers eight customer-service domains and 112 grounded cases, conducts multi-round dialogues with a trust-tracking customer agent, and separates deception emerging from incentive alone from deception produced under explicit instruction. Across eighteen proprietary and open-weight models, emergent deception varies substantially across model families and domains. We further use the benchmark for post-training, finding that honesty-directed fine-tuning reduces deception under incentive, while deception-graded fine-tuning increases lie success on honest-control dialogues without increasing lie frequency under incentive. By verifying entitlement knowledge before scoring deceptive behavior, KnownLieBench reduces the confound between lying and not knowing and enables more rigorous auditing and steering of agent honesty.

Large language models (LLMs) exhibit uneven multilingual performance, especially when dealing with low-resource languages. Inference-time intervention offers a lightweight way to improve cross-lingual transfer by modifying the hidden states produced by the LLMs during the forward pass, without updating model parameters. However, existing cross-lingual intervention methods typically learn separate projections from source to target languages, which limits scalability and prevents knowledge sharing across languages. We propose Centroid Intervention Fusion (CIF), a projection fusion framework that consolidates multiple multilingual intervention projections into a single language-shared operator. Across multilingual commonsense reasoning, natural language inference, factual editing, and machine translation benchmarks, CIF outperforms the strongest prior pairwise intervention baseline by up to +3.378 pp on average across four model backbones, while supporting performance gains for low resource languages. The code is available at https://github.com/VRCMF/CIF.git.

While LVLMs rapidly improve, long-video question answering still remains challenging: relevant evidence is sparse, and question-relevant context often fails to provide cues that discriminate the correct answer from plausible alternatives. Diagnostic analysis on a manually annotated subset of MMR-V shows that prior agentic systems substantially improve cue retrieval over direct VLM inference yet fail to achieve a corresponding gain in answer accuracy, indicating that the bottleneck lies in option-discriminative evidence rather than topical relevance alone. We propose PACE (Progressive Acquisition of Critical Evidence), a factor-guided framework for long-video evidence acquisition. PACE proceeds in two stages: it first indexes clip-level descriptions guided by question-derived factors without observing the candidate answers; it then uses the candidate answers to derive contrastive cues and queries the index for verification. On MMR-V with the open-source Qwen3-VL backbone, PACE achieves 42.6% accuracy, outperforming direct inference and prior agentic baselines including Deep Video Discovery (DVD). On the same diagnostic subset, PACE recovers 66.9% of the annotated cues, providing empirical evidence that its gains are associated with improved evidence recovery rather than stronger answer-side priors alone. Consistent gains over DVD on LVBench, Video-MME, EgoSchema, and LongVideoBench suggest that option-aware evidence acquisition transfers beyond MMR-V. Code is available at https://github.com/HKUST-KnowComp/PACE.

Simulation-based inference (SBI) for parameter estimation is vulnerable to model misspecification: neural summaries and density estimators trained on a specific forward model typically fail when applied to data drawn from another model, or from real observations, and no training simulator can capture the full observational pipeline of a real measurement exactly. We show that a self-supervised Vision Transformer (ViT), pretrained label-free on a fast approximate simulator, produces transferable data summaries that generalize across simulators. Without retraining, it can be reused as a frozen encoder to infer astrophysical parameters from a completely different simulator that resolves the radiative transfer explicitly, on which it has never seen either data or parameters. As a concrete use case in 21cm cosmology, SKATR, a ViT pretrained with a Joint Embedding Predictive Architecture (JEPA), serves as a foundation model for reionization inference from upcoming SKA measurements: SKATR is pretrained once on 67k low-cost, noiseless semi-numerical 21cmFAST lightcones, then frozen and applied to hydrodynamical Loreli II lightcones, where a lightweight conditional flow matching head infers five astrophysical parameters; the encoder is never shown Loreli data, its parameters, or any noise. In our comparison, SKATR yields the most precise and best-calibrated posteriors across all five parameters, matching the accuracy of the fully-supervised in-domain baseline while requiring 2.6x fewer radiative-transfer simulations. Under realistic SKA AA* noise, only SKATR remains simultaneously accurate, informative, and calibrated, outperforming even a supervised baseline retrained from scratch on noisy data. Self-supervised pretraining on computationally efficient semi-numerical simulations is therefore a viable route to calibrated, simulator- and noise-agnostic reionization inference for the SKA-era.

We investigate whether AI-generated impulsive sounds can be distinguished from real ones through group delay analysis. Our central finding is that AI-generated impulsive sounds show near-identical onset-region group-delay distributions but exhibit measurably different group-delay behavior in the late decay region: decay-region KL divergence reaches $0.322$ compared to near-zero onset divergence ($0.022$). Cross-band GD variability achieves single-feature AUC~=~0.720, and a Random Forest (RF) over nine decay-region features reaches AUC~$=$~0.884 under sample-disjoint evaluation. A group delay map used as a standalone 2D input to CNN classifiers achieves 90--94\% accuracy, demonstrating that group delay carries substantial discriminative information. Under generator hold-out, CNN and transformer classifiers show highly variable AUC (0.457--0.918). The group delay RF achieves the highest average hold-out accuracy among the evaluated methods ($66.7\%$) and avoids extreme below-random collapse, although its average AUC (0.731) is lower than CNN avg (0.762) and AST (0.772). Parameter sensitivity analysis across 27 STFT configurations confirms that the RF AUC remains stable (0.700--0.847, std~=~0.035). These results suggest that decay-region group delay can serve as a physically interpretable forensic cue that complements magnitude-based classifiers, while broader validation remains necessary.

Spreadsheet formulas can refer to rectangular ranges of arbitrary size. When a user changes the structure of a referenced table, the spreadsheet system updates the references to refer to a new range. Unfortunately, this new range may differ from the user's expectations, introducing bugs in spreadsheets. We describe a user study showing that standard reference semantics are error-prone, resulting in significant risk to users. We introduce Kale, a prototype system that eliminates the risk of inserting these kinds of bugs by restricting the kinds of references that can be expressed. We show that Kale can be used effectively by users to complete tasks that are error-prone in traditional spreadsheet systems. Finally, we describe a corpus study that evaluates the extent to which the reference restrictions in Kale might have implications on users.

We previously reported a ModernBERT encoder trained from scratch for Turkish (MoganBERT-TR) and a single-vector embedding model built on top of it (MoganBERT-embed). This work introduces the third model in that lineage: MoganColBERT-TR, a multi-vector retrieval model that, instead of compressing a query or a document into a single vector, represents it at the token level through a 768->128 projection and scores it with MaxSim late interaction. The model is not trained from scratch: the embedding model's encoder is taken as the starting point and adapted to the ColBERT objective with a single-epoch distillation phase. Training data is produced from two sources - title-to-passage pairs carved out of our own pretraining corpus in the character domain and at sentence boundaries, and two Turkish question-based retrieval sets - and is distilled from the soft scores of a cross-encoder teacher (bge-reranker-v2-m3) over one positive and seven mined negatives. We show that in hard negative mining, rank-based skipping alone is insufficient and must be combined with a group mask and a cosine ceiling. Evaluation is carried out with the official pipeline of TurkColBERT, a benchmark built for Turkish late-interaction retrieval (PLAID index, exact MaxSim), on five Turkish BEIR datasets; none of them appears in our training pool, so all five results are clean zero-shot. With 148.9M parameters, MoganColBERT-TR reaches an overall score of 37.36 (35.53 nDCG@100, 31.81 nDCG@10) averaged over the five datasets and finishes second among the five models compared: it outperforms the twice-as-large ColmmBERT-base-TR on four of five datasets and by +3.05 overall, and the benchmark's largest model by +12.30. The gap to the leading model (mLateOn) is concentrated on ArguAna-TR, the dataset with by far the longest queries.

Automated theorem proving offers a natural foundation for recursive self-improvement in scientific discovery. However, existing neural provers do not fully preserve this recursive structure, where the learning process should be self-improving over time. Existing methods either embed proof experience into model parameters through expensive weight updates, or keep verified intermediate deductions only within the current problem. In addition, these methods also heavily rely on sparse whole-proof feedback, even when unsuccessful partial attempts contain useful discoveries. To close the gap, we propose ProofEvolve, a neuro-symbolic framework that evolves explicit, formally verified symbolic proof structures with neural models to decisively expand the knowledge boundary. In this framework, the neural model proposes variation operators, including decompositions, repairs, and schema recombinations. The symbolic Lean kernel verifies every proof transition. Over the evolution loops, ProofEvolve computes verified closure over the resulting proof directed acyclic graphs (DAGs). Within each problem, ProofEvolve evolves partial AND-OR proof DAGs in a behaviorally indexed archive. Across problems, kernel-checked schema extraction adds newly proved sub-DAGs to a persistent schema library. Proof DAGs inherit the solved results through typed schema recombination, with every residual premise exposed as a new subgoal. This evolutionary process preserves verified results from incomplete attempts and makes them available for later proofs without weakening formal soundness. Across three competition-level Lean benchmarks, ProofEvolve achieves the highest average solve rate among the evaluated proof systems.

Managed LLM services are now part of real production systems, but model selection and service planning still rely heavily on capability benchmarks that reveal little about operational behavior after deployment. We present Operational Embedding (OpEmbed), a framework for learning compact operational fingerprints of LLM cloud services from structured, privacy-preserving support-case metadata, without using case text. OpEmbed aggregates model--time windows into an eight-channel operational signature and learns a low-dimensional representation via temporal contrastive learning, cross-view reconstruction, and generational-ordinality regularization. Evaluated on more than 33,000 production support cases spanning seven LLM families over 26 months at Google Cloud, OpEmbed recovers interpretable family- and version-level structure, improves leave-one-model-out operational forecasting over non-learned baselines, remains useful under limited early-window data, and supports cross-model fault-type transfer. We report the practical lessons learned from building and evaluating this tool for model onboarding, support readiness assessment, and operational monitoring.

While tool-augmented Large Language Models have significantly improved multi-step reasoning in quantitative STEM tasks, a critical residual failure mode remains: intermediate reasoning steps that are syntactically well-formed, mathematically executable, and unit-consistent, yet contextually ungrounded. Current approaches either rely on formal verifiers that cannot assess semantic intent, or burden Process Reward Models (PRMs) with the dual task of checking both arithmetic and logic. In this paper, we propose a neuro-symbolic framework that cleanly decouples reasoning into two formal dimensions: Symbolic Validity ($V$) and Semantic Groundedness ($G$). We guarantee $V$ by construction using a deterministic symbolic verifier acting as a hard filter. To assess $G$, we train a PRM conditionally on the verifier-accepted manifold. To train this PRM efficiently, we introduce Counterfactual Symbolic Perturbation (CSP), a novel data synthesis strategy that algorithmically generates constraint-preserving hard negatives (steps that perfectly pass the verifier but are logically flawed). At inference, we deploy a verifier-first constrained search that guarantees execution consistency for verifier-covered operations while relying on the PRM solely to rank semantic grounding. By targeting the exact residual error class of strong tool-using LLMs, our method significantly improves reasoning reliability without the sprawling heuristics of prior frameworks.

Large language models increasingly produce and interpret verbal probability expressions, yet whether these expressions carry consistent meaning across models (or match human perceptions of uncertainty) remains unknown. We present a systematic cross-model evaluation using a word-to-number mapping task grounded in established human benchmarks. Eleven uncertainty expressions were presented to 19 models under two conditions, forced single-number response and explanation elicitation, alongside a novel bidirectional roundtrip test of internal consistency. LLMs track the human benchmark with surprising fidelity: word ordering is preserved, three anchor points are recovered, and ``possible'' shows the highest variance and cross-model disagreement of any expression tested, consistent with its documented bimodal interpretation in humans. However, models show a systematic upward bias for negative expressions such as ``unlikely'' and ``improbable.'' Explanation elicitation reduces within-model variance while increasing between-model divergence, stabilizing individual models at the cost of inter-model consensus, and the roundtrip experiment reveals clear stratification, with frontier models maintaining coherent bidirectional representations. LLMs thus reproduce the structure of human verbal probability cognition, including its biases, while diverging systematically at the negative end---with implications for any setting where humans and models exchange probabilistic language.

Best-of-N (BoN) sampling is the simplest and most widely deployed inference-time alignment strategy, but it suffers from two distinct problems: reward hacking, in which the selected response exploits errors in the proxy reward model, and the absence of any privacy protection for the sensitive human preference data used to train that reward model. We show that a single intervention-adding calibrated noise to reward scores before selection-resolves both. Our first result, Private Best-of-N (PrivBoN), establishes that Gumbel noise at an appropriate scale simultaneously provides $ε$-differential privacy and implements KL-regularized alignment. Whenever the privacy budget exceeds a critical threshold $ε^*$, the privacy-mandated noise is the regret-optimal regularization, and privacy imposes zero additional alignment cost-matching the information-theoretic skyline of Huang et al. (2025). Because $ε^*$ depends on an unknown coverage coefficient, we introduce Private Inference-Time Pessimism (PrivITP), which combines $χ^2$-regularized rejection sampling with a two-phase Gaussian mechanism. PrivITP achieves ex-post $(ε,δ)$-DP with a privacy cost independent of the number of responses $n$, cleanly decouples the regularization parameter from the privacy parameter, and attains the skyline up to a noise-inflation term. Experiments across several language models, datasets, and reward models confirm our results: PrivBoN and PrivITP are scaling-monotonic (unlike BoN, which degrades past a critical $n$), and PrivITP matches or outperforms PrivBoN at equivalent privacy levels, with the largest gains in the strong-privacy regime.

The performance of textual neural models often degrades when their inputs are corrupted by noise such as typos, OCR errors, or dropped words. We study the degradation rate across neural models, both sentence embeddings and decoder-only LLMs, and find that how consistent it is depends on the scale of the noise: under word-level noise, models with very different architectures decline along nearly the same curve, while under character-level noise they separate. We further identify the determining factor to be the training objective, not the architecture: eight encoders spanning six pretraining paradigms are scattered initially, and collapse onto a common curve after a short contrastive training recipe. We trace the word/character split to tokenization: a single character edit forces the tokenizer to re-segment the surrounding word, disturbing the token sequence far more than dropping a whole word does. This finding and its underlying mechanism provide a practical means to predict a model's robustness to noise without any noisy evaluation, and to install robustness at a chosen noise scale through noise-augmented training.

Frontier language models are increasingly marketed as omni systems that can perceive and respond across modalities. Existing evaluation frameworks, however, focus almost exclusively on bimodal understanding, typically text plus one other modality. We propose the Modality Maturity Index (MMI), a benchmark designed to evaluate the multimodal capabilities of large language models across five modalities (text, image, audio, video and document) and combinations of up to three modalities in both inputs and outputs. MMI consists of 893 questions, each carefully crafted to require the model to demonstrate its understanding of multiple input modalities and to generate responses that incorporate various output formats. The questions are designed to be self-contained, with clear expectations for the correct modality or mix of modalities required for an accurate response. Every MMI prompt carries human-authored rubric criteria for each output modality expected in the response; a model's MMI Value expresses the average of the per-modality scores for each prompt. Because low scores can reflect either failure to generate a modality (lack of presence) or failure to generate correct content, we introduce also a supplementary Modality Presence Score (MPS), a per-prompt F1 over the expected output modalities. Applying MMI to five frontier multimodal models, we find that the MPS ranges from only 15.6 (Claude Opus 4.6) to 34.9 (GPT-5.4). Given the low availability of returned modalities to even grade, we report MPS as our main result pending model improvements. To assess the viability of judging output correctness with LLM judges and rubrics, we run a separate experiment with custom generation tools. On the assets that generates, we find that an LLM judge applying the rubrics agrees with rubric-blind human annotators (who score the outputs directly and never see the criteria) on 70.8% of judgments.

AI-assisted development tools enable software engineers to generate implementations at substantially higher speed and volume than in traditional workflows. Software teams have long relied on guardrails -- standing control mechanisms such as code review, linting, testing, and CI/CD pipelines -- to maintain quality and coordination. High-throughput AI-assisted generation increases pressure on these guardrails -- straining their capacity to keep pace with the volume and rate of generated changes -- and reshapes how organizations supervise development workflows, yet relatively little is known about how existing guardrails evolve in response. We conducted a qualitative interview study with five software engineering practitioners, situated within a broader practitioner survey. Our findings indicate that organizations distribute the work of supervision across multiple guardrail layers: preventive guardrails (produced by externalizing architectural intent and conventions into machine-interpretable form), executable guardrails (linting, testing, and CI/CD repurposed as scalable supervision infrastructure), and human oversight (shifting from line-by-line inspection toward supervisory interpretation focused on architectural reasoning, explainability, and long-term maintainability). We characterize this as a transition from review- centric guardrails toward layered supervision, in which no single guardrail carries the supervision load alone.

Large language models can now generate complex, multi-step mathematical proofs, but reliably determining their correctness and localizing early logical errors remains a critical challenge. Existing evaluation approaches largely depend on model-based natural-language judgments, which often overlook local reasoning gaps. While formal theorem provers like Lean offer a path to rigorous verification, using them to evaluate informal text requires solving locality and semantic mismatches: a prover might bypass a local flaw by proving an overly broad target, or validate an auto-formalized statement that drifts from the original mathematical intent. To address this, we introduce FaithSieve, a Lean-assisted framework for fine-grained evaluation of natural-language mathematical proofs. FaithSieve decomposes coarse proof steps into local reasoning units, extracts typed proof obligations, and verifies them through a formal evaluation agent. Formal validation is gated by semantic alignment scoring, so Lean evidence is incorporated only when the formal statement faithfully preserves the context, objects, and logical form of the original claim. We construct two expert-verified datasets, ProofLoc-Olympiad and ProofLoc-University, to benchmark first-error localization. On the 350-problem Olympiad dataset, FaithSieve using a GPT-5.4 backbone achieves 81.43% exact first-error accuracy, outperforming the direct-judging baseline of 72.29%. Furthermore, on the 200-problem ProofLoc-University benchmark spanning six advanced domains, FaithSieve reaches 84.5% exact accuracy, compared to 75.0% for the direct judge. Our work demonstrates that decomposing proofs into fine-grained units and grounding them with faithful formal evidence significantly improves reliable evaluation of natural-language reasoning.

Modern machine learning models rely on large amounts of labeled data. However, manual annotation of large-scale datasets is expensive and time-consuming. Label spreading is a semi-supervised learning technique that addresses this challenge by propagating information from a few labeled examples to a larger pool of unlabeled data. Despite its effectiveness, its application to large-scale, high-dimensional datasets is limited by computational costs and memory constraints. To address these limitations, we propose Algebraic Multigrid Acceleration for Efficient Label Spreading (AMELS), an efficient label spreading framework that improves scalability by fast construction of neighborhood graphs and the incorporation of algebraic multigrid solvers. The latter is an iterative solver that replaces the ordinary random walk iteration typically performed in label spreading. Due to the multilevel nature of algebraic multigrid solvers, AMELS spreads given label information across a graph of any size in a single multigrid cycle. We demonstrate that AMELS achieves significant runtime reductions compared to existing implementations while also being more robust to hyperparameter choices in terms of both runtime and classification accuracy. Our framework therefore enables efficient label spreading on large-scale image datasets and produces accurate labels even when only a few labeled samples are available.

A large language model (LLM) guardrail for a self-adaptive system (SAS) may issue an approval that is correct at check time but stale by actuation. This creates an Execute-stage time-of-check to time-of-use (TOCTOU) hazard. We study verdict freshness: whether a guardrail verdict remains valid when used. We distinguish three quantities that answer different questions: all-candidate verdict change under fixed-action replay, oracle-labeled approval expiry on recorded closed-loop trajectories, and judge-conditioned use-time invalidity. Across five reproducible SAS environments, all-candidate verdict-change rates span 5.3-48.4% at a common replay shift of eight simulator steps. We introduce the Freshness-Bounded Shield (FBS), which estimates each approval's validity horizon from its safe-side margin and recent feature volatility, without an explicit plant-dynamics model. Using fixed settings documented in the artifact, FBS reduces oracle-labeled approval-expiry rates from 3.4-24.7% to 0-1.8% at the same shift. A separate audit of four LLM judges finds nonzero judge-conditioned use-time invalidity in every approval stream. We formulate a freshness contract: every approval must be correct at check time and remain valid at use time.

Urban food environments may reflect broader socioeconomic inequalities, but district-level evidence remains limited in Brazilian cities. This study examined whether indicators of food retail and street-market availability discriminate between levels of social vulnerability across the 96 districts of São Paulo. We conducted an exploratory cross-sectional ecological analysis integrating the São Paulo Social Vulnerability Index (IPVS), establishment records from the Relação Anual de Informações Sociais (RAIS), and street-market data from CAISAN. Census-sector information was aggregated at the district level. Twenty districts without an IPVS classification were excluded, resulting in 76 observations. The outcome distinguished districts classified as IPVS level 1 from those classified as levels 2--7. Predictors described the densities of healthy and unhealthy food establishments, the number of street markets, and the availability of establishments selling fresh or in natura food. Eight conventional machine-learning classifiers were evaluated using leave-one-out cross-validation. Reported mean F-scores ranged from 0.62 to 0.75, with XGBoost obtaining the highest value. In the Random Forest model, the densities of healthy and unhealthy food establishments jointly accounted for approximately 60% of the total impurity-based feature importance. These findings indicate that publicly available food-environment indicators contain information associated with the district-level distribution of social vulnerability. However, the small ecological sample, class imbalance, outcome binarization, and cross-sectional design limit predictive generalization and preclude causal or household-level interpretations.

Tool-augmented LLMs must arbitrate between two fallible sources when a tool return conflicts with their parametric memory, yet existing evaluations measure source preference without establishing source correctness. We introduce MemToC, a controlled benchmark for post-tool-return arbitration with executable tools. MemToC comprises 6,504 evaluation episodes constructed from 542 quality-controlled factual questions, independently elicited model-specific closed-book answers, and controlled tool returns of known correctness. These components instantiate four source-correctness cases; tool-error and no-tool conditions are separate controls. Across five open-weight 7-9B models, tool returns strongly dominate elicited closed-book answers. The four instruction-tuned models retain a verified-correct answer against an incorrect tool in only 6.5-17.1% of eligible cases, follow a correct tool in 86.0-93.1%, and repeat the tool return in 78.4-86.0% of cases where both sources are wrong. No cross-model ordering remains stable across three instruction-wording variants with the question and episode content held fixed. We compare prompting with SFT and DPO using chain-level cross-fitting over ToolHop, so questions sharing an underlying fact never straddle training and evaluation. We apply an asymmetric success criterion: correct-answer retention must improve without a detected reduction in correct-tool following. SFT and DPO meet this criterion on the same two of four instruction-tuned backbones. Improvements rarely come cleanly: 19 of 20 tested method-model combinations reduce abstention after tool errors or on unanswerable inputs. Transfer beyond MemToC is positive but partial and depends on the model and presentation frame. Correctness-conditioned arbitration can be improved through fine-tuning, but gains must be evaluated jointly with correct tool use, abstention, and robustness to formulation.

Every memory-based knowledge editor in the SERAC lineage depends on a scope decision: given a query, does a stored edit apply? We report that current knowledge-editing benchmarks cannot measure this decision at all. Using INLAY, a gradient-free editor we built to obtain exact per-query ground truth (the model is frozen, edits live in an external addressable memory, and applying an edit is a bias added along one token's unembedding direction at decode time), we execute every candidate router action on 1,689 queries spanning three datasets and three input conditions. An oracle router choosing the best action every time ties a one-line static policy to four decimal places in all nine dataset-by-condition cells: the maximum attainable gain of any per-query routing method is 0.00 points. Abstention is the sole winning action zero times out of 1,689. The cause is structural: these are counterfactual benchmarks whose evaluation question asks for the post-edit answer, so answering from parametric knowledge is wrong by construction, and a benchmark without negatives cannot reward a classifier's ability to reject. This generalizes beyond our system to the whole scope-classifier family the benchmarks are used to evaluate. We confirm the mechanism directly: constructing the missing condition ourselves, by withholding a query's own edit from the index for half the sample, moves pooled headroom from exactly +0.0000 to +0.0420 and gives abstention its first wins. We also report where INLAY itself does not win (WISE beats it on Qwen2.5-7B CounterFact, and retrieval-augmented generation beats every method we tested, INLAY included, on rigorously matched RippleEdits), and disclose two bugs found during a self-audit of our own routing machinery, neither of which changed a published headline number outside noise.

Mentalization - the ability to infer others' beliefs and intentions to guide one's own choices - is a key cognitive function underlying human social interactions. Large language models (LLMs) demonstrate behaviour consistent with humans on theory-of-mind tasks, yet whether these models can guide adaptive behaviour through mentalization is unknown. Here we use two economic games with cognitive computational modeling to uncover the latent strategies underlying mentalization in LLMs. We tested individual LLM agents across four model families, DeepSeek, GPT-4.1, GPT-5 and Gemini 2.0 Flash (N = 2,099), against opponents of varying sophistication and examined whether a prompting strategy designed to elicit strategic reasoning improved performance. We benchmarked results against human participants (N = 251) as a comparative measure. Across both games, LLMs showed clear behavioural and computational signatures of mentalizing that differed markedly by model provider and size. Strategic prompting generally improved performance by inducing more sophisticated reasoning, yet the extent of the benefit differed across the two tasks. Last, GPT-5 agents flexibly adapted their recursive depth of reasoning to increasingly sophisticated opponents, demonstrating superior performance to human participants. Collectively, we demonstrate different capacities for mentalization across LLMs, and highlight cognitive computational modeling as a formal method for assessing comparative intelligence across humans and machines.

Muon has emerged as a strong optimizer for the matrix-valued parameters in large language model pretraining, approximately orthogonalizing its momentum with a few Newton-Schulz iterations. Existing theory either replaces this iteration with the exact polar factor it approximates, or treats its finite depth as an approximation error, and thus the iteration Muon actually runs can only hurt the guarantees. We show that finite Newton-Schulz can instead be beneficial for nonsmooth nonconvex optimization. To this end, we analyze Muon through the online-to-nonconvex conversion, which views the update rule as an online learner and converts its regret bound into a stationarity guarantee. The finite Newton-Schulz iteration smooths the discontinuous polar map into a Lipschitz map of the singular values, and Muon with finite Newton-Schulz can be regarded as an online learner with a smoothed spectral potential. This smoothing is exactly what the conversion needs: we prove that a Newton-Schulz depth growing only logarithmically in the target accuracy suffices for convergence to stationary points in nonsmooth nonconvex optimization, whereas Muon with the exact-polar update may fail to converge. The resulting sample complexity bounds match the best-known guarantees for nonsmooth nonconvex optimization and are optimal for smooth nonconvex optimization up to problem-dependent factors. The argument extends beyond Newton-Schulz to general spectral maps with the same smoothing property.

Extracting a shared set of unknown, not directly measurable quantities from multiple, heterogeneous datasets is a common challenge across scientific domains. A prominent example is the combination of datasets obtained from different measurements with different settings (e.g. varying detector resolutions). Analyzing such datasets jointly, rather than independently or after naive merging, is essential for obtaining precise and unbiased estimates of the unknowns, but requires careful treatment of dataset heterogeneity and is computationally demanding. We present a generalized framework for simultaneously analyzing multiple heterogeneous datasets in the context of generative AI-based inverse problem solvers. Building on our recent Scalable Asynchronous Generative Inverse Problem Solver (SAGIPS) framework, we extend the well-established distributed data-parallel training paradigm to non-identically distributed datasets, where each dataset is controlled by the same set of unknown inference parameters but covers a different region of the available feature space. Each dataset is processed through its own forward operator and discriminator, providing complementary constraints that collectively guide a shared generator toward global parameter consistency. We validate the approach using a controlled setup inspired by a multi-detector scattering experiment. We provide numerical evidence that our framework is robust to different data fidelities, which arise from unknown detector systematics in the Rutherford experiment, and we show the scaling behavior on multi-GPU leadership computing systems. The results show that our approach is well suited for real-world multi-dataset analyses in which experimental conditions vary across measurements.

Static shape estimation of co-manipulative continuum robots (CCRs) is challenging because the continuum arms and manipulated flexible object form a closed chain that must satisfy both static equilibrium and geometric loop-closure constraints. This paper presents a constraint-aware physics-informed neural network (PINN) for static shape estimation of a tendon-driven CCR modeled using the geometric variable strain formulation. The proposed method incorporates a projected static equilibrium residual and a configuration-level geometric residual to enforce the governing mechanics and closed-chain geometry. In simulation, the PINN is compared with a purely data-driven artificial neural network (ANN) under limited and noisy training data. With 140 samples and 50% label noise, the PINN reduces the relative configuration error, equilibrium residual, and closed-chain residual by 67.88%, 67.35%, and 88.06%, respectively. Using the full dataset, the PINN achieves 0.1597% relative configuration error with an inference time of 0.1773 ms, compared with 17.97 s for an iterative nonlinear solver. Experimental fine-tuning reduces the marker RMSE from 2.657 mm to 0.497 mm and increases R2 from -0.788 to 0.937. These results demonstrate accurate, physically consistent, and computationally efficient static shape estimation of closed-chain CCRs.

Large Language Models (LLMs) increasingly act as autonomous agents executing complex, long-running procedural skills. Existing agent runtimes maintain execution by continually appending observations, actions, and intermediate reasoning traces to an ever-growing conversation history, causing latency degradation and context-poisoning failures over long horizons. We present SKILL.state, a runtime architecture that replaces append-only conversational history with an explicit, mutable execution state. At each execution step, the model receives only the immutable skill specification, the current structured execution state, and the latest observation. Intermediate reasoning is discarded immediately after producing a validated state update, preventing prompt growth with execution history. Across diverse datasets, models, and execution environments, SKILL.state improves task accuracy while substantially reducing cumulative token consumption. Our results demonstrate that explicit execution state is an effective and architecture-agnostic abstraction for scalable long-horizon agent skills.

Capture-the-Flag (CTF) benchmarks are widely used to assess the offensive security capabilities of autonomous language-model agents. Evaluations rely on shallow binary judgments or aggregate scores, overlooking the agent's trajectory to the flag. Consequently actual exploitation is conflated with direct flag exposure, memorized recall, external lookup, guessing, and unsupported claims, potentially overstating the agent's cybersecurity capability. We introduce CTF-ABACUS, a trace-based agent auditing framework that reconstructs each run as an evidence-grounded solve profile. By decomposing agent actions into penetration-testing phases and categorical techniques, it identifies where exploitation occurs, where the flag first appears, and whether the recovered flag is supported by demonstrated behavior. Aggregating solve profiles across agents yields challenge signatures that reveal whether success was achieved via the intended exploit or via shortcut pathways. We apply CTF-ABACUS to 1,435 CTF attempts by six frontier and open-source models on 240 challenges, yielding 2,870 solve profiles under two judge lenses. Trace-verified exploits account for only 62-87% of recovered flags across benchmarks, while shortcut recoveries follow substantially shallower trajectories. These findings shift CTF evaluation from counting recovered flags to verifying demonstrated exploitation and provide a basis for designing benchmarks that better isolate the offensive capabilities.

We present a six-stage framework for auditing the reproducibility of scientific claims across a research literature within the computer science domain, and instantiate our framework for the neuro-symbolic AI (NSAI) subdomain. Instantiating the framework on the NSAI subdomain produced a multi-year audit. Stage one retrieved 5,497 records and removed 3,018 duplicates. Stage two screened the 2,479 unique records at title and abstract, identifying 1,365 self-identified NSAI records, then removed a further 61 at full text for off-topic, non-research, no-quantitative-evaluation, or inaccessible-full-text reasons. Stage three sought a verifiable public code artifact for each of the 1,304 eligible records and found none for 849, leaving 455 to enter the artifact inventory and bounded rerun of stages four and five. We fully or partially reproduced 85 studies, 6.52% of the eligible corpus and 18.68% of attempted reruns. We found that 321 attempted reruns were blocked by missing non- code artifacts and 42 by missing or unusable code repositories. These figures quantify a persistent reproducibility deficit that survives even nominal "code available" declarations, and signal the need for enforced, versioned, and permanently archived artifact bundles in future NSAI publications. We argue that empirical NSAI papers should be required at submission time to provide complete, versioned, and permanently archived artifact bundles.

Accuracy-only benchmarking of reasoning-capable large language models misses a central deployment question: when do extended thinking tokens earn their cost? We introduce the Token Economy Score (TES), a marginal benchmarking metric that measures the accuracy gain of a reasoning model over a non-reasoning baseline, normalized by the generated-token multiplier. We define paired and approximated TES variants for model families with reasoning toggles and frontier models without direct non-reasoning counterparts. We then conduct an empirical benchmarking analysis across 151 model-benchmark evaluation runs on seven benchmarks spanning mathematics, code generation, science reasoning, instruction following, expert knowledge, knowledge recall, and research-level physics. The analysis examines three deployment-facing dimensions: which task structures yield positive marginal reasoning efficiency, how increasing reasoning effort changes TES within model families, and how deployment context changes economic viability. Results show that task structure predicts reasoning efficiency better than nominal difficulty: sequential inferencechain tasks such as AIME 2025 and LiveCodeBench show high TES, while knowledge-recall tasks such as MMLU-Pro show low TES despite their difficulty. We also find systematic diminishing returns at higher reasoning effort levels, including cases where additional thinking reduces accuracy. Finally, Reasoning Cost Share (RCS) shows that inference spend is often dominated by internal thinking, while Deployment Cost Multiplier (DCM) shows how on-premises deployment can change the economics of otherwise costly reasoning workloads. These findings support a benchmarking-driven model-selection rule: enable reasoning selectively by task type, effort level, and deployment context rather than treating it as a universally beneficial mode.

Retail electricity markets in deregulated systems face significant price volatility and complex interactions with forward and futures products, posing challenges for effective operational decision-making. This study introduces a Causal Graph-Informed Temporal Convolutional Network (CG-TCN), a forecasting architecture that integrates a learned causal graph into a temporal convolutional network via a graph-neural embedding to enhance both forecasting accuracy and interpretability of retail electricity price dynamics. It first applies a multi-resolution decomposition to isolate semiannual, quarterly, and monthly trends from high-frequency fluctuations. A causal graph is then discovered over these components and key covariates, including wholesale forward prices and retail contract attributes such as early termination fees, with domain constraints that preserve causal directionality and exogeneity. The learned causal structure is encoded as an adjacency embedding that conditions the TCN's convolutions and attention, aligning representation learning with causal pathways. Using ten years of daily 12-month fixed-price residential contracts from Ohio's deregulated market, we find that wholesale forward prices primarily determine long-term retail price trends, whereas contract attributes influence short-term fluctuations. CG-TCN consistently outperforms benchmark models, achieving mean absolute percentage errors of 3.08%, 3.82%, and 5.43% for one-, ten-, and fifteen-step-ahead forecasts of daily retail electricity median prices, respectively. By combining predictive performance with interpretability, CG-TCN provides transparent, policy-relevant insight to support market analytics, consumer protection, regulatory oversight, risk assessment and procurement planning in competitive electricity markets.

Extreme compression of deep neural networks, up to full binarization, dramatically reduces memory footprint and arithmetic complexity, facilitating deployment on constrained edge hardware with field-programmable gate arrays (FPGAs) and microcontrollers. Although combining binarization with pruning promises additional efficiency gains, existing pruning strategies are ill-suited to binarized representations and rarely translate into meaningful hardware savings. We introduce a PyTorch-based, research-oriented framework that incorporates freezing and pruning mechanisms for designing and optimizing binarized neural networks. The framework enables rapid and reproducible evaluation of state-of-the-art approaches and the fast prototyping of new ones. Leveraging this framework, we propose a novel pruning method that accounts for the relative importance of learned parameters across abstraction levels. Such a global weighting mechanism consistently achieves a superior trade-off between model accuracy and pruning rate, achieving a 70% pruning rate on VGG11 with constant accuracy, while state-of-the-art results reach only 41% in the binarized setting.

Large-kernel Convolutional Neural Networks (CNNs) deliver remarkable performance in vision tasks by significantly expanding receptive fields, yet their quadratic parameter growth critically impedes storage-efficient edge deployment. While existing efficient architectures adopt parameter-efficient depthwise separable convolution backbones that leverage techniques like low-rank approximation and weight sharing to compress depthwise convolutions, we identify a critical oversight: pointwise convolutions dominate parameter volume (>87% in models like RepLKNet-31B) and constitute the primary deployment bottleneck on resource-constrained edge devices. This results in prohibitive storage costs and severe memory-loading constraints on resource-limited devices (e.g., smartphones with 4-12 GB Random Access Memory (RAM)). To overcome this, we propose Channel Group-Shared (CGS) low-rank approximation, a novel Singular Value Decomposition (SVD)-based parameter-sharing strategy. CGS constructs a structured low-rank paradigm isomorphic to SVD decomposition, comprising shared (high-parameter-cost) down/up-projection matrices across channel groups within a layer and channel-group-specific (low-parameter-cost) scalable diagonal matrices. This group-sharing design achieves significant parameter reduction. Extensive experiments demonstrate that large-kernel CNNs (RepLKNet, ConvNeXt, SLaK) enhanced with CGS strike an empirically favorable balance between competitive performance and substantially reduced storage costs. Crucially, by alleviating storage constraints, reducing memory bandwidth pressure during loading, and minimizing model loading latency, CGS enables the feasible deployment of pre-trained large-kernel CNN models on edge devices, thereby bridging the gap between high-performance vision models and practical edge deployment.

Organizations assign documentation work to technical writers, yet the knowledge required to produce it is distributed across developers, managers, and other practitioners. Prior work has established quality criteria for judging "good" documentation, but it has not examined how practitioners bring that expertise to improve documentation quality or the challenges they face in doing so. Through semi-structured interviews with experienced technical writers ($n=31$) from different organizations, our work reveals the individual and collaborative effort required to maintain documentation quality. We identify five distinct stages of the documentation review process: self review, technical review, editorial review, play testing, and post-publication feedback. Each stage draws on practitioners with distinct expertise to address quality across content, presentation, and user experience. Our findings surface organizational and technical challenges writers face in recruiting expert reviewers, navigating development timelines, and contending with tools not specifically designed for documentation workflows. Our work positions documentation review as a crucial yet understudied site of collaborative work and opens new research and design directions for process improvement and tool support.

Multimodal instruction-following models require training data that is accurate, diverse, verifiable, and challenging. Existing synthesis pipelines typically follow a one-pass generate-and-filter paradigm, discarding feedback from failed samples, verifier outcomes, and target-model errors. We present VISA (Visual Instruction Synthesis Agent), an agentic framework that reformulates multimodal instruction synthesis as a self-evolving loop. At each round, VISA analyzes an image to filter incompatible constraints and discover new verifiable ones, samples diversity- and difficulty-aware constraint sets from persistent memory, generates candidate instructions, and verifies the resulting samples with executable tools and structured large language model judges. Failed samples trigger diagnostic-guided recovery, while accepted samples are probed against the target model to estimate difficulty. The resulting verifier signals and target-model failure profiles are written back to memory, allowing subsequent rounds to adaptively expand the constraint space, reduce template repetition, and focus on unresolved model weaknesses. The same verifier contracts further provide reward signals for reinforcement learning without a separately trained reward model. Experiments on MM-IFEval show that VISA consistently improves multimodal instruction following over strong baselines, while preserving general multimodal capability across seven public benchmarks.

Energy consumption and climate change have made sustainability critical in Software Engineering (SE), driving the emergence of Green SE. Over the past 15 years, numerous solutions for sustainable software systems have been published by the SE community, offering a rich resource for analyzing the field's evolution. To explore this, we conducted a systematic mapping study of Green SE research published between 2010 and 2024. We collected 390 publications, categorizing them by application domain (e.g., mobile, cloud, AI) and research type (e.g., optimisation study, benchmarking, literature review Additionally, we analyzed a representative subset of 79 papers to classify the key elements-such as hardware, measurement, stability, and replicability-considered during energy measurement experiments. Our findings indicate that SE conferences host the majority of energy-related literature. Notably, Green SE studies surged in popularity starting in 2023, largely driven by AI-related publications. Optimization and benchmarking emerged as the most prevalent research types. Ultimately, we aim to inform the SE community about current approaches to energy concerns, highlight critical experimental practices, and advocate for continued action toward more sustainable software engineering.

The rapid advancement of Large Language Models (LLMs) makes it increasingly difficult to distinguish human writing from machine-generated text. Training-free detection offers a scalable solution, yet common confidence-based metrics mainly measure average token probabilities and often miss the signal fluctuations that characterize human writing, which we call "generative vitality". Spectral analysis offers a way to capture this vitality, but its mechanism and practical boundaries remain underexplored. In this paper, we analyze spectral detection from both theoretical and empirical perspectives. We connect spectral energy to variance in proxy log-probability trajectories and explain how broader human token choices create the fluctuations used by frequency-domain indicators. We further show that the strength of this signal depends on text length and sampling range: spectral evidence is clearest for long, continuous, constrained generation, while short, fragmented, mixed, and edited settings require complementary confidence and fluctuation views. These findings clarify when frequency-domain detection works and provide guidance for future multi-dimensional detector design.

LLM-based agents are increasingly being developed for time-series problems, but their design choices vary substantially across task settings. This survey adopts a problem-driven taxonomy that organizes these systems by the time-series problems they address rather than by isolated technical components. We group existing systems into four categories: forecasting and reasoning, augmentation and synthesis, anomaly detection and diagnosis, and decision support. Within each category, we examine how task requirements shape agent architecture, tool use, and memory design. We further summarize representative datasets and environments, and compare reported model performance under shared or closely related settings. Overall, this survey offers a task-oriented guide to designing LLM-based agents for time-series problems and identifies open gaps for future work.

Autonomous agents increasingly perform bounded software tasks under an orchestrator that retries, resumes, and budgets them. The machinery such orchestrators reach for is the service mesh's: retry, timeout, and error-rate circuit breaking. We report a failure study of a production agentic software-delivery platform over 147 numbered incidents spanning 81 runs, each with a measured cost and, in most cases, a mutation proof reproducing the failure. All three assumptions those primitives rest on are violated in practice, and we quantify the consequences: a loop of fifty-four consecutive successful tool calls no error-rate breaker could see; a progress signal constant by construction, guaranteeing a false trip on the third repair round and driving one run from six of six components to three; twenty-one events accumulated across six invocations of one delegation, making a correct, idempotent component unwinnable; a misrouted failure that woke five components for a two-component fault, leaving three bystanders regressing working code; and twelve incidents in which the enforcement layer blocked correct work, the most expensive costing 107 agent turns and zero accepted writes. We find one cross-cutting cause and its dual. Identity adequacy: in five separate subsystems an identity that failed to discriminate produced a confident wrong answer, and two of them derived the corrective rule independently. Evidence adequacy: a reliability decision may be taken only on evidence capable of moving, attributable to what it measures, and deterministic under identical conditions. From the findings we derive seven reliability primitives whose enforcement unit is the delegation rather than the message, and specify the controlled evaluation the study motivates but does not constitute.

Event triggering sits at the heart of high-energy physics, where the rare events of interest must be retained while an overwhelming background is discarded under tight latency and bandwidth budgets. This work compares four classical machine learning models, namely a support vector machine, an artificial neural network, a convolutional network and a long short-term memory network, with four hybrid quantum counterparts, on a trigger-like binary classification task built from CMS open data. The label is defined by an invariant-mass window, and the inputs combine reconstructed kinematics with physics-motivated derived variables: the pseudorapidity difference, the wrapped azimuthal difference, the angular separation and the total transverse momentum. The quantum models run under a fixed resource budget of eight qubits, a principal-component compression to sixteen features and state-vector simulation. Every model shares the same stratified split, the same preprocessing and a common decision threshold, and performance is reported through accuracy, ROC-AUC, F1-score, precision and recall. The strongest classical model is the artificial neural network, at 93.53 percent accuracy and 0.9819 ROC-AUC, while the strongest quantum model is the quantum convolutional network, at 90.89 percent accuracy and 0.9731 ROC-AUC, with the quantum neural network close behind. The quantum-kernel and recurrent quantum approaches trail both, which places the trainable hybrid embeddings ahead within this budget. The study is meant as a controlled reference point rather than a claim of quantum advantage.

Generalist vision--language--action (VLA) policies learn long-horizon behavior mainly through short-horizon action prediction and reveal little beyond sampled commands. This creates two coupled bottlenecks: a single action target must implicitly absorb task progress, intermediate intent, and local reliability, while these control states remain hidden during execution. Inspired by functional principles of biological sensorimotor control, we introduce LM-X , which organizes prediction across task, event, and motor scales without claiming anatomical correspondence. Three explicitly supervised signals are emitted online and directly condition action generation: return-to-go (RTG) measures visible task progress, event-to-go (ETG) identifies the next semantic transition, and heteroscedastic action flow estimates local reliability through propagated variance. Explanation is therefore intrinsic to control rather than generated post hoc. Before a costly 20-day pretraining run on 64 NVIDIA B200 GPUs, a controlled five-task pretraining gate verifies the design: the complete model improves success by 16.0 points over the action-only backbone and by 10.8 points over the strongest single-head variant. We then train LM-X on more than 20,000 hours of real-robot trajectories, including over 1,000 hours of failed policy rollouts. LM-X achieves 74.1\% across 50 randomized-hard RoboTwin2.0 tasks versus 55.4\% for GR00T N1.7, and 68.6\% versus 50.7\% across seven real-robot tasks. RTG tracks semantic progress and visible regression, while variance rises during hesitation and oscillatory control. These results show that explicit multi-timescale predictive state can strengthen control while exposing interpretable internal estimates.

Safety evaluation is critical for assessing whether aligned Large Language Models (LLMs) remain robust against jailbreak attacks. Existing automated testing methods, however, largely rely on response-level feedback: each candidate prompt typically requires generating a target-model response to evaluate its attack effectiveness. This process is expensive and, more importantly, provides only sparse guidance on strongly aligned models, where most candidates are rejected with the same failure outcome. This paper presents NeuronFuzz, a white-box fuzzing framework that exploits internal safety neurons as continuous execution feedback for LLM safety evaluation. A SafetyOracle converts safety-neuron activations into a continuous safety alarm score that serves as feedback for fuzzing and can be obtained during prefill, eliminating response generation from the fuzzing loop. To construct the SafetyOracle, NeuronFuzz uses template-invariant harmful and benign inputs and stability-aware selection to identify a compact set of safety neurons whose activations capture harmful-intent recognition. Moreover, since the safety alarm score is differentiable, NeuronFuzz uses its gradients to identify safety-sensitive template positions and a masked language model to generate fluent, context-compatible mutations while preserving original harmful payload and avoiding additional optimization variables. We evaluate NeuronFuzz across 21 text and multimodal models. Across five white-box source models, it achieves a 76-100% jailbreak discovery rate, outperforming baselines by up to 48 percentage points. Its optimized templates further transfer zero-shot to open-weight and six proprietary target models, achieving average ASR and top-5 ensemble ASR (EASR) of 69.6%/92.6% and 44.1%/60.0%, respectively.

As generative AI gains traction, researchers are investigating its potential to serve as proxies for humans. From undergoing cognitive psychology experiments to experiencing an epidemic, generative agents, agents powered by generative AI models, produce realistic human behavior when prompted. This study explores the sensitivity of these generative agents' behavior to prompt modifications and varied persona names of the agents. To assess this sensitivity, we use a generative agent epidemic model, wherein each agent is prompted daily on whether it wants to isolate or commingle with other agents. We found that using synonymous prompts results in negligible changes to the model's outcomes. However, minor variations in prompts, as well as contextual changes, do influence the model's results. Lastly, our data indicates that different persona names assigned to generative agents, specifically those imbued with personas, do not significantly impact epidemic outcomes.

Reconstructing continuous physical fields from sparse measurements is central to scientific monitoring, inverse modeling, and digital-twin construction. Generative reconstruction has recently emerged as a promising paradigm for this task by learning data-driven physical priors that complete plausible full fields from limited observations. However, existing methods largely assume fixed, batch conditioning, whereas real sensing systems often produce structured streams: probes scan local regions, instruments observe moving fields of view, and communication constraints may leave entire frames missing. We propose TRACE, a retrospective streaming generative reconstruction framework for physical fields under structured sensing. TRACE performs approximate Bayesian inference in a learned continuous-coordinate latent space, converting sparse off-grid measurements into generative latent evidence, fusing it with a state-space temporal prior through Kalman-style filtering, and refining under-observed past frames via retrospective smoothing. Experiments on active matter, ocean sound-speed fields, and supernova simulations show that TRACE matches or surpasses frame-wise generative reconstructors, offline spatiotemporal methods, and streaming data-assimilation baselines in reconstruction quality under temporally sparse and spatially localized sensing protocols.

A coding agent combines a model with a harness, which decides what the model sees, which tools it can use, and how the work continues. We ask whether changing the harness changes the result when the model and task stay fixed. We compare two configurations of the same harness on three coding benchmarks. The control supplies the full conversation in time order, while the treatment keeps the same record but mechanically shortens older tool results as the context fills and responds to repeated or stalled work. Under tight context, the treatment raises mean per-task fail-to-pass fraction (F2PF) in all three pressure comparisons and increases complete solutions on SWE-bench Verified and SWE-bench Pro. The tight-window Verified comparison uses 169 tasks, a 20,480-token window, and a fixed 480-second attempt endpoint; on this cohort, treatment raises mean per-task F2PF from 28 percent to 49 percent and complete solutions from 43 to 72. Without model-specific retuning, the same frozen treatment also raises both endpoints on the same cohort for three additional models with different designs. In the wide-window Qwen3.6 comparisons, observed arm outcomes are close on Verified and Pro, while FeatureBench retains a higher mean per-task F2PF under treatment. On the wide-window Verified cohort, treatment also serves fewer prompt tokens per turn. Because changing the harness changed what unchanged model weights could accomplish, coding-agent evaluations should treat the model and harness together as the tested solver.

Parameter-efficient fine-tuning is often assumed to preserve pretrained capabilities because it updates only a small number of parameters. We show that this assumption depends strongly on adapter capacity. We study factual acquisition in a controlled OpenStreetMap-derived benchmark where Qwen3-4B must acquire anonymized geographic associations while retaining unrelated capabilities. Comparing full fine-tuning (FFT) with quantized low-rank adaptation (QLoRA) at ranks 8, 16, 32, and 64, we find that rank induces a clear acquisition--retention frontier. Low-rank QLoRA preserves out-of-domain (OOD) performance but acquires fewer facts, whereas higher ranks improve same-fact paraphrase generalization at an increasing cost in performance on unrelated benchmarks. FFT behaves as a conservative baseline: it retains general capabilities well, but does not reach the highest factual-acquisition regime. Distributional, weight-space, and spectral diagnostics mirror this behavioral trade-off, with higher-rank QLoRA moving farther from the pretrained model. A separate math adaptation experiment shows a weaker frontier, suggesting that the effect is most pronounced when adaptation must install new factual associations rather than reinforce skills already supported by pretraining. Code and data are available at https://github.com/zhngstl/new_facts_forgetting.

On-policy distillation (OPD) supervises a student on its own trajectories with token-level signals from a frozen teacher, yet how a sampled loss allocates updates across tokens remains poorly understood. We analyze the gradient of the per-token K2 estimator of reverse KL with respect to the student logits. The $\ell_1$ norm of this gradient factorizes into the absolute teacher--student log-probability gap and a student-side softmax factor that grows as the sampled token becomes less likely under the student. In our math-distillation runs, these per-token norms are highly non-uniform: low-student-probability tokens account for a disproportionate share of their sum and are also enriched in large teacher--student gaps. As a lightweight intervention suggested by this analysis, we study Surprise-aware Reweighting (SuRe), a detached, bounded weighting rule that further amplifies this existing allocation. Across two Qwen3 student scales, SuRe improves several math metrics over vanilla OPD and shows no clear degradation on the selected out-of-domain benchmarks. Our primary contribution is therefore a gradient-level characterization of reverse-KL OPD trained with the K2 estimator, with SuRe as one empirical instantiation.

The deployment of advanced, AI-enabled control algorithms in tokamak experiments requires robust integration with existing plasma control system (PCS) architectures and extensive pre-experimental validation. In this contribution, we describe the integration and testing of neural-network-emulated virtual circuits for plasma shape control within the MAST Upgrade (MAST-U) PCS environment. The neural network models predict the plasma shape using the plasma current, poloidal field coil currents, and plasma profile parameters. In this paper, we explain how they are deployed via a real-time C++ inference server that interfaces with the PCS, returning the shape prediction and its Jacobian, and how, from the latter, virtual circuit matrices and updated coil current requests are computed for real-time actuation. Emphasis is placed on the validation workflow and best practices adopted to ensure confidence in the proposed control framework prior to experimental deployment. This work demonstrates practical AI-based shape control components for fusion control systems, with direct relevance for upcoming MAST-U experiments and future devices.

Artificial Intelligence (AI) is increasingly embedded in modern software systems, raising important questions about how its quality should be defined, assessed, and assured. This paper presents a Systematic Mapping Study (SMS) on the quality of AI-based software. The study synthesizes primary studies published between January 2020 and January 2026 and selected from five electronic data sources. A total of 33 primary studies were included after automated search, screening, and snowballing. The results identify six recurring challenge categories, with the most prominent being limitations in existing quality assessment models, followed by issues in non-functional requirement management, quality-aware development, and quality assurance. The findings suggest a call for collaboration of researchers and industrial practitioners with standardization organizations, that could possibly devise comprehensive quality assessments and their measurement methods.

An agent that inherits a consolidated memory may inherit a constraint that was true when written and has since been withdrawn by a newer authoritative record. Under a scarce verification budget, does the agent recover the withdrawal, and if not, is the resulting stale-consistent decision avoidable without spending more? We model supersession explicitly -- provenance is immutable; what changes is which record is current -- and assign by design the memory's form, the world's state and the verification policy at a fixed budget of two records: the agent's own allocation, or the same budget with one slot re-assigned to the critical provenance path or to a random record. With a constraint stated, agents inspected its provenance path in about one episode in five; when that constraint had been superseded, native allocation produced stale-consistent decisions in 77.3%, 74.7% and 74.7% of episodes across a primary run, a fresh-wording replication and a held-out domain. Re-assigning one slot to the critical path raised current-record-consistent decisions by +74.0, +72.7 and +61.3 points, positive in six of six models in each run, and left an already near-ceiling rate unchanged when the record agreed with the memory. The held-out scenario was later found to contain a temporal inconsistency; a robustness replication with one sentence corrected, deposited externally before execution, gave +73.3 points (positive in 5 of six models, the sixth at a native missed-path rate of zero) and is reported alongside the original. The intervention uses knowledge of the critical path and is not a scheduler; it quantifies how much of the stale-consistent decision rate the bundled same-budget policy removes: the effect approaches the native missed-path rate in the primary, replication and corrected held-out runs. Memory systems may need freshness or supersession signals separate from relevance.