Artificial Intelligence
SkillAlchemy: Open-World Agent Skill Creation
Hengjun Wang, Shuyue Wei, Boyi Liu et al. (+2 more)
Agent skills are reusable procedural artifacts that extend
language agents with specialized workflows, tool conventions, and domain behaviors at inference time. However, creating reliable skills still depends largely on human authorship, model priors, or execution traces. These sources are often unavailable for unfamiliar tasks, suggesting the need to create skills from open-world materials. In this paper, we study open-world skill creation: given an underspecified skill brief and a source-access specification, a creator must discover behavior-relevant requirements omitted by the brief and determine how broadly each source-derived procedure is justified. We propose SkillAlchemy, an admission-centered framework for source-grounded skill creation. SkillAlchemy identifies implicit requirements through contrastive evidence, admits candidate procedures based on evidence-supported scope, and compiles the admitted content into a grammar-guided skill package. Extensive experiments across 87 SkillsBench v1.1 tasks demonstrate that SkillAlchemy improves pass rate over no-skill execution by 19.9 percentage points and the strongest automated baseline by 8.6 percentage points, while achieving performance comparable to human-curated skills.
Read the full abstract ▾
All authors Hengjun Wang, Shuyue Wei, Boyi Liu, Jun Yang, Yongxin Tong
Artificial Intelligence · Cross-listed from cs.IR
Adaptive Item-based Collaborative Structures via Noise Rescheduling in Diffusion for Generative Recommendation
Jiaqi Wang, Tianying Liu, Heng Chang et al. (+3 more)
Discrete
Diffusion Models (DDMs) have recently been introduced to recommendation systems, modeling user history as a
token generation process via iterative denoising. However, while effective at capturing user-level sequential patterns, these methods often fail to explicitly integrate item-based collaborative filtering information, a critical component for accurate recommendation. This deficiency manifests in two key aspects: (1) the item representation is often semantic-focused, lacking collaborative priors for
diffusion training; and (2) the denoising process employs a uniform noise schedule, treating all
tokens indiscriminately and ignoring item-level adaptive structural dependencies. To bridge this gap, we propose ANR-DiffRec, a unified framework designed to encode item-based collaborative structures into discrete
diffusion for generative recommendation. First, we explicitly incorporate an item co-occurrence
matrix to guide semantic ID generation, providing a structured collaborative prior for discrete
diffusion training. Second, we introduce an item-based adaptive noise rescheduling mechanism that dynamically adjusts denoising weights according to both local contextual recoverability and behavior-aware item dependencies. Specifically, the proposed strategy jointly models intra-item structural context and inter-item collaborative signals, enabling structure-aware denoising during
diffusion training. Extensive experiments on multiple
benchmarks demonstrate that our method consistently outperforms state-of-the-art generative recommendation models. Code: https://github.com/CalmaQi/ANR-DiffRec.
Read the full abstract ▾
All authors Jiaqi Wang, Tianying Liu, Heng Chang, Jihong Guan, Wengen Li, Shuigeng Zhou
Artificial Intelligence
MediSkill-Evo: Process-Constrained Self-Evolution for Evidence-Grounded Clinical Interaction
Ruoyu Wu, Shenfu Xie, Yinqian Sun et al. (+2 more)
Interactive clinical
agents operate under partial observability, so reliable care depends on reaching the correct diagnosis through evidence-grounded, safe interactions. Yet existing
agents struggle to convert experience into reusable process knowledge with explicit provenance and authority. To address this gap, we introduce MediSkill-Evo, which self-evolves governed process knowledge without fine-tuning the backbone. It realizes this self-
evolution by updating clinical, process, symbolic, and visual knowledge in four typed banks under type-specific validation and scope rules. The Process-Constrained Preference Harness then turns validated knowledge into action by grounding candidates in evidence and prioritizing safer decisions. We evaluate on 300 MIMIC-IV-derived FullChain encounters, 180 hard-isolation conditions covering six process obligations, and 100
multimodal NEJM image-diagnosis cases. On Qwen FullChain, MediSkill-Evo improves diagnosis accuracy by 7.81% and treatment-intent coverage by 70.67% over the best-performing prior
agent, while reducing critical failures by 43.04%. Under stress, it improves the stress-process composite by 7.77% and required-action completion by 12.41% over the best-performing
agent for each metric, with stronger patient-fact, temporal-evidence, and triage-red-flag recovery and no controller-scored errors in unavailable-evidence, treatment, and triage safety checks. On
multimodal NEJM diagnosis, MediSkill-Evo with optional MedSAM localization improves diagnosis accuracy by 2.56% and core score by 18.96% over the best-performing
memory agent. Code is available at https://anonymous.4open.science/r/mediskill-evo_anonymous-68E7.
Read the full abstract ▾
All authors Ruoyu Wu, Shenfu Xie, Yinqian Sun, Haibo Tong, Feifei Zhao
Artificial Intelligence · Cross-listed from cs.MA
Right-Sizing LLM-Agent Decomposition in VAT Determination: A Pilot Controlled Sweep
Pedro Santos
Recent LLM-
agent systems make conflicting design bets: decompose work across many narrow
agents, or use one strong tool-using
agent. This pilot studies that choice on bounded cross-border VAT determination with reverse charge, where every case has an oracle label and each intermediate decision is independently scoreable. We hold the activity surface fixed (subtasks, tools, I/O schemas, validation checks, orchestrator, base model, and merge policy) and vary only the assignment of subtasks to workers across four orchestrated configurations, from one wide worker to five narrow ones, against S0, a tuned no-orchestrator single
agent, with a deterministic rule engine as oracle. The program spans 4,400 runs: a 40-case, five-repeat main sweep, matched-
token arms separating
prompt-budget from
agent-count effects, and three failure-injection arms, all judged against pre-registered falsification criteria. The two intermediate configurations lead on accuracy (0.830, against endpoints at 0.720 and 0.770) but miss the pre-stated bar against the fine endpoint, so the intermediate-optimum hypothesis remains unsupported at pilot scale. The single
agent does not Pareto-dominate the orchestrated set. The matched-
token criterion fires: the budget-matched single
agent lands 6.5 points below the leader, but the interval includes zero, so any advantage is consistent with a
prompt-budget explanation. Under injection, availability faults are absorbed at every granularity, with wide-scope restart over-recovering its baseline by +0.160, while one schema-conforming hallucinated record degrades every configuration and inverts the ordering, hitting fragmented configurations hardest. The contribution is a bounded, preregistered pilot heuristic for right-sizing decomposition (place one partition boundary at the dependency-layer midpoint), released with oracle,
dataset, harness, raw traces, and analysis pipeline.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.IR
Towards a Densing Law for User Representation Learning at Billion-Scale Capacity
Bin Dou, Junru Zhang, Zhaoyi Yuan et al. (+6 more)
User representation learning in real-world industrial scenarios is commonly scaled by increasing user amount, behavioral sequence length and model size. However, existing methods face two challenges: (i) Bottleneck for raw data scaling at billion-scale capacity, as performance exhibit diminishing performance gains with larger-scale raw text user behavioral input, which can be mitigated by tokenization. (ii) Lack of quantitative analysis of how tokenization configurations should scale with data size. In this report, we propose User Behavioral Densing Law for characterizing the quantitative relationship between data scale and the minimum sufficient tokenization capacity. Firstly, we conduct a pilot study on raw & tokenized scaling comparison on billion-scale Alipay
dataset, revealing the raw data scaling bottleneck and the sustained gains enabled by tokenization. To derive the scaling pattern governing the minimum sufficient tokenization configuration at different data scales, theoretical analysis and systematic experiments are employed to summarize the quantitative scaling pattern. We find an approximately linear relationship between the logarithms of minimum sufficient tokenization capacity and input data size measured by
tokens, and the scaling slope varies systematically with the tokenization method and data source, reflecting differences in representation-space redundancy and intra-source uniqueness. Guided by the proposed law, we further develop ALGN, an adaptive variable-length tokenization method that improves capacity allocation. Extensive experiments across diverse data sources, tokenization methods, and downstream tasks demonstrate the generalizability and reliability of the User Behavioral Densing Law, providing practical guidance for tokenization configuration selection in large-scale user representation learning. Moreover, ALGN outperforms existing baselines.
Read the full abstract ▾
All authors Bin Dou, Junru Zhang, Zhaoyi Yuan, Wuliang Huang, Letian Gong, Baokun Wang, Huan Li, Yu Cheng, Weiqiang Wang
Artificial Intelligence · Cross-listed from cs.CL
Cross-Domain, Multi-Task Data-to-Text Generation without In-Domain Training Data
Yifei Song, Kun Efimov-Zhang and Claire Gardent
Structured data exists in many forms (tables, knowledge graphs, charts, and time series), and converting it into text may involve different generation tasks. However, most prior work on data-to-text (D2T) generation has focused on specific tasks and
datasets, relying either on task-specific training data or on the zero-shot capabilities of large
language models. We study cross-domain D2T generation in a setting where neither in-domain training text nor test references are available, and where domains, generation goals, and input structures vary substantially. We compare data-driven knowledge
distillation (DDKD) against zero-shot inference and fine-tuning on out-of-domain D2T data, and introduce structure-preserving augmentation via structural subsampling and perturbation. Experiments on five
benchmarks show that, at constant model size (1.7B parameters), DDKD consistently outperforms both fine-tuning and zero-shot inference. Moreover, the resulting small models outperform a much larger finetuned model on two of the five domains, achieving comparable performance on the remaining three. We further construct QUINTD-5, a fivefold extension of QUINTD-1, and show that simply scaling real target-domain inputs yields only modest gains, whereas our augmentation strategy remains more effective and more cost-efficient for cross-domain
distillation.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.CL
Cross-lingual Biography Enrichment via Claim Extraction and Alignment
Yifei Song, Ziyang Chen, Emil Sayilov et al. (+1 more)
English Wikipedia is often treated as the default encyclopedic source, yet non-English Wikipedia editions can contain richer locally grounded information for long-tail figures. We study cross-lingual biography enrichment: enriching an existing English biography with facts supported by a non-English biography about the same person. Focusing on women from non-English-speaking contexts, we introduce CLAW-4L, a
benchmark consisting of 300 Wikipedia biography pairs linking an English biography with its French, Chinese or Azerbaijani counterpart, along with claim annotations and a fine-grained claim-pair relation corpus. We propose a claim-based enrichment framework that extracts English claims from both biographies, aligns them to identify enrichment evidence from the non-English biography, and rewrites the English biography using the selected claims. Our results show that non-English Wikipedia biographies provide valuable evidence for improving English biography coverage, while lower-resource settings remain challenging.
Read the full abstract ▾
All authors Yifei Song, Ziyang Chen, Emil Sayilov, Claire Gardent
Artificial Intelligence · Cross-listed from cs.CR
Adversarial Entropy Inflation Against Gumbel-Based Inference Verification
Nikita Kezins
Gumbel-based inference verification bounds LLM weight exfiltration by only forgiving
token choices that plausibly arise from honest GPU nondeterminism, reporting a >200x slowdown for a steganographic adversary under benign
prompt traffic. This bound assumes a passive attacker; we show it degrades sharply against an adversary who instead controls the
prompt distribution. Because the verifier's admissible-
token-set size is driven by the model's own output
entropy,
prompts engineered to break grammatical and sub-word structure – rather than benign conversational traffic – widen that set and open a materially larger covert channel. Across six instruction-tuned models spanning 1B to 32B parameters and three random seeds, our strongest attack (character- and script-level disruption) roughly doubles bits leaked per
token relative to benign
prompts, cutting the slowdown factor to 60x - 118x. These results indicate that static, benign-traffic-calibrated thresholds are insufficient for this defense, and that jitter-forgiveness thresholds should instead be calibrated dynamically against local
token entropy.
Read the full abstract ▾
Artificial Intelligence
Modalities Should Talk to Each Other: Dual-Stream Multimodal Learning for Long-Horizon Influenza Forecasting
Seyed Mohammad Hossein Hashemi, Mohsen Hooshmand and Parvin Razzaghi
Forecasting long-range influenza-like illness (ILI) matters for public health readiness. Publicly available surveillance
datasets typically pair numeric epidemiological signals with textual information that is noisy, loosely structured, only indirectly related to near-term trends, and often lagged relative to the numeric signal. Fusing the two therefore requires careful design. We propose Dual-Stream
Attention (DSA), a
multimodal deep learning framework that forecasts 12-week-ahead ILI activity from a 36-week
multimodal history by letting the numerical and textual streams condition each other. Using the Time-MMD health-domain
dataset, DSA separately encodes the two modalities with a
Transformer-based numerical encoder and a domain-adapted headline encoder, then couples them through a bidirectional Cross-Modal
Attention (CMA) mechanism: the text (news headlines) conditions the interpretation of the numeric signal and vice versa. The CMA output then passes to a causal temporal model for
forecasting. Evaluated across ten random seeds, DSA achieves a median test MSE of 0.416, versus 0.668, 0.607, and 0.851 for iTransformer, TaTS, and GPT4MTS, corresponding to mean-error reductions of 54.95%, 37.29%, and 67.23%, with paired Cohen's d of 0.555, 0.337, and 0.345, respectively, and ranks first in 100% of bootstrap draws. It also has substantially lower worst-window error than all baselines. On an external-geography
dataset, DSA again ranks first among nine evaluated baselines. Ablations show the advantage does not depend on text-encoder choice or
language-model fine-tuning, and that bidirectional
attention outperforms either direction alone. Finally, perturbation-based faithfulness analysis shows the learned CMA is functionally informative under targeted masking, with a stronger effect in the text-to-numerical direction.
Read the full abstract ▾
Artificial Intelligence
Walking on the DARKSIDE
Aldo Gangemi and Emanuele Bottazzi
Large
Language Models (LLMs) do not natively track the path of exclusions that a coherent discourse demands. When an input rests on a fabricated authority, a misapplied mechanism, or a surreptitious analogy, an unsteered LLM tends to engage with it as if it were well-posed, and this affects its generation. POLANYI++, an LLM-steering method that uses heuristics, ontologies and problem-solving methods for tacit-knowledge extraction, produces an Extended Knowledge Graph (XKG) in OWL2, but when a sophisticated nonsensical input is reified into the graph alongside the legitimate triples, it gets hardly detectable by automated reasoners, since the XKG is generated jointly with the wrong assumptions. We introduce DARKSIDE, a coherence-
auditing method on top of POLANYI++. DARKSIDE formalises an explicit data structure of accumulated exclusions over discourse time, complemented by a warrant axis that classifies each named referent as Warranted, Unattested, Misattributed or Fabricated. The method is anchored in nine theoretical fragments unified under a shared deep frame of path integrity. The resulting DARKPOLANYI is evaluated as a steering layer over Gemini 3 on BSBench, a 100-item
adversarial corpus of sophisticated-sounding nonsense across multiple domains, with Claude Sonnet 4.6 as an independent judge. DARKPOLANYI scores 1.89/2 mean versus 0.95/2 for the unsteered Gemini 3 Pro baseline; on the 97 cases with valid judgments in both arms, paired McNemar gives a paired bootstrap mean-diff = +0.92 (95% CI [+0.75, +1.08], p = 0.0001). The evidence supports an architectural claim: when an LLM forward pass is wrapped in an ontology-mediated
auditing, structurally inevitable
hallucination can be partially recovered. The XKG functions as the missing
memory that LLMs lack, and the warrant axis as an epistemic firewall.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.CV
DF-MoE: Generalizable Deepfake Detection via Multimodal Sparse Mixture-of-Experts
Vlad Hondru, Florinel Alin Croitoru, Iuliana Georgescu et al. (+2 more)
Audio-visual deepfake detection is an actively studied topic, where one of the main challenges is to develop detectors able to generalize across deepfake generation methods. We
conjecture that overfitting can be mitigated by extracting multiple high-level cues from the available audio and visual modalities via pre-trained models. We therefore assemble a wide variety of pre-trained models to extract features that encode mouth movements, face parsing, facial expressions, head pose, gaze tracking, heart rate, audio
emotion and
speech activity. We further integrate both unimodal and
multimodal cues via a Mixture-of-Experts (MoE) backbone to detect deepfakes. We perform in-domain and cross-domain experiments on five
benchmarks for deepfake detection (MAVOS-DD, AVLips, PolyGlotFake, BioDeepAV, FakeAVCeleb) to compare our framework (DF-MoE) with state-of-the-art methods. Our results indicate that DF-MoE obtains superior deepfake detection results, surpassing all competing methods. We release our code at https://github.com/vladhondru25/DF-MoE.
Read the full abstract ▾
All authors Vlad Hondru, Florinel Alin Croitoru, Iuliana Georgescu, A. Sophia Koepke, Radu Tudor Ionescu
Artificial Intelligence · Cross-listed from cs.CL
The Emergence of Relevance Through Axiomatic Attention Patterns During LoRA Fine-Tuning
Matthew Perlman, Atharva Nijasure and James Allan
LoRA fine-tuning is standard for adapting LLMs to reranking, but it remains unclear where in the network task-specific relevance behavior is learned and what
attention-level changes accompany that learning. Through ablation and
attention experiments, we identify where LoRA
attention updates to RankLLaMA improve performance and whether those gains coincide with interpretable relevance-oriented
attention patterns such as lexical matching, rarity sensitivity, and query-document interaction. We find that given LoRA fine-tuned MLPs throughout the network, restricting LoRA
attention updates to a compact mid-network region is sufficient for recovering over half of the performance gained by applying LoRA to all
attention layers, and that omitting
attention fine-tuning in this region hurts performance more than elsewhere in the network. Additionally, we show that regions where applying LoRA affects performance the most overlap with regions where fine-tuning increased
attention to axiomatic IR features. Rarity sensitivity, document-query interaction, and several compositional features are highly correlated with gains in ranking performance. Our results support an interpretable, correlational account of how relevance-oriented behavior emerges during LoRA fine-tuning and point toward improved strategies for adapting rerankers.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.CV
Thinking Beyond Videos: Unifying Video Reasoning and Deep Research for Open-World Video Agents
Wenqi Liu, Shijie Ma, Yunxiao Wang et al. (+21 more)
Open-world
video understanding often requires a model to locate sparse visual evidence and acquire external knowledge that is absent from the
video and its parametric
memory. While Thinking-with-
Videos enables active temporal perception and Deep Research supports multi-step information seeking, the two capabilities are typically developed in isolation. We introduce VideoRover, a unified
Video Deep Research framework that iteratively coordinates
video cropping,
multimodal search, and webpage browsing. Given a
video-question pair, VideoRover uses each tool result to select the next action, so localized
video clips guide external
retrieval and retrieved evidence triggers further
video inspection and verification. To develop this capability, we construct an automated data curation pipeline, producing 26K verified SFT trajectories and 3K challenging RL instances. We also introduce VideoRover-Bench, a
benchmark stratified by
video duration and research difficulty. Experiments on VideoDR and VideoRover-Bench show that our VideoRover-8B-RL achieves performance comparable to proprietary models in the direct-answer setting without tool use while outperforming larger open-source models equipped with the same tool suite. Ablation studies and training dynamics further validate the complementary roles of active
video grounding, external
retrieval, and long-horizon reinforcement learning.
Read the full abstract ▾
All authors Wenqi Liu, Shijie Ma, Yunxiao Wang, Meng Liu, Qile Su, Han Liu, Bohan Hou, Zeyu Wang, Xuanyu Zheng, Changyi Liu, Tianke Zhang, Haonan Fan, Kaiyu Jiang, Yingxin Li, Jiankang Chen, Xu Wang, Hongyi Fu, Jianxiong Wang, Bin Wen, Tingting Gao, Han Li, Jianhua Yin, Yinwei Wei, Xuemeng Song
Artificial Intelligence · Cross-listed from cs.NE
Mycelial Search: A Graph-Structured Metaheuristic for Continuous Optimisation
Mohammad Mahdi Dehshibi
Continuous optimisation methods need to balance sharing information and maintaining alternative search directions. In this paper, we introduce Mycelial Search (Myco), a graph-structured metaheuristic designed around active tips, community-weighted flow, adaptive cord plasticity, and anchor-based injection. Candidate solutions form an evolving spatial graph in which a Louvain partition distinguishes within-community from cross-community information exchange. Adaptive cord plasticity subsequently modifies active tip-to-tip edges according to their
alignment with the local flow. An anchor-based injection mechanism supplements the graph-driven tip dynamics. We evaluated Myco on the CEC 2022 single-objective bound-constrained
benchmark suite at dimensions $D=10$ and $D=20$, using 30 independent runs per
algorithm-function pair. The comparison includes eleven established optimisers from several search families. Myco reaches competitive results on selected functions across both dimensions. The ablation analysis further shows that community structure regulates the range of graph-based information exchange, whereas cord plasticity controls the persistence of local directional influence. These findings indicate that graph-structured local interaction can support continuous optimisation, while its effectiveness depends on landscape structure and information transfer across local search regions.
Read the full abstract ▾
Artificial Intelligence
Agent-G$^2$: Gaussian Guidance for Agentic Reinforcement Learning
Zixuan Wang, Yanrui Miao, Zhengxi Lu et al. (+6 more)
Hint-based reinforcement learning addresses reward sparsity in long-horizon agentic tasks by retaining a prefix of an expert trajectory before each rollout, letting the policy explore from a state closer to success. Its effectiveness hinges on the guidance depth: how much of the trajectory to keep. Existing methods treat this depth as a deterministic scalar. Scheduled approaches share one value across samples and ignore per-task heterogeneity; per-sample probing estimates it separately at the cost of extra rollouts. We find that useful guidance occupies a band of depths whose informativeness profile is approximately Gaussian around the band center, rather than concentrating at a single optimal point. We propose
Agent-G$^2$, a Gaussian guidance framework that draws the depth per task from a Gaussian whose center and spread are estimated online from rollouts already collected for policy optimization, requiring no probe rollouts or learned depth predictor. The center combines a global baseline with per-cluster difficulty, and the spread tracks within-cluster variance. We evaluate
Agent-G$^2$ on ALFWorld and WebShop on Qwen2.5-1.5B / 7B-Instruct.
Agent-G$^2$ outperforms the strongest hint-based, hint-free, and Aux-RL baselines on ALFWorld by 2.3 / 3.9 / 7.4 points at under one-third the rollout cost of per-sample probing.
Read the full abstract ▾
All authors Zixuan Wang, Yanrui Miao, Zhengxi Lu, Teng Pan, Yiwen Qiu, Hongxing Li, Peng Qiu, Ruiqing Zhang, Yongliang Shen
Artificial Intelligence
EviSafe: Evidence-Grounded Safety Evaluation for Vision-Language Models
Xuetong Li and Gaofeng Liu
Vision-
language model safety
benchmarks typically evaluate only final responses: whether a model refuses, warns, or complies. This outcome-level view cannot tell whether a model is safe for the right
multimodal reason. Safelooking behavior may reflect keyword-triggered refusal, missed visual hazards, or over-refusal of benign-sensitive inputs. We introduce EviSafe, an evidence-grounded framework for VLM safety that jointly evaluates natural user-facing behavior, explicit grounding in textual and visual evidence, and behavioral sensitivity to counterfactual changes in safety-critical evidence. EviSafeBench instantiates the framework as a controlled
benchmark with 1,181 gold image-text scenarios and 2,452 targeted counterfactual variants across eight safety domains and eight risk-source types. Each scenario includes a gold safety decision, evidence annotations, a safe-response policy, and counterfactual interventions. The three-probe protocol queries models with natural-response, evidencereporting, and counterfactual-response
prompts, then scores them using an evidence-aware judge. Across eleven evaluated VLMs, natural severity accuracy ranges from 27.6% to 52.8%, relaxed diagnostic consistency from 6.1% to 29.3%, and unsafe-to-safe counterfactual transition success from 30.4% to 58.4%. These gaps show that the evaluated VLMs are not reliably safe for the right
multimodal reason and motivate evaluation beyond refusal counts.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.CR
FIDES: A Concordance Protocol for LLM-Generated Trading Strategies
Arther Tian, Alex Ding, Simon Wu et al. (+1 more)
An LLM asked for a trading strategy returns three artifacts at once: a natural-
language rationale, an executable implementation, and once run, a track record. Whether these are the same object is rarely checked. We present FIDES, a measurement protocol that treats them as three views to be reconciled rather than one deliverable to be graded. Through dual delivery, a single model call returns both a natural-
language strategy with an explicit claimed edge and a self-contained strategy(df) function. FIDES executes the code in a sandbox against a lag-one out-of-sample backtest and scores three concordance gaps: say to do, do to real, and say to result. On 8
liquid US ETFs across four models plus a two-stage elicitation arm, 40 strategies, 2023 to 2024 out-of-sample, three findings stand out. First, concordance does not predict profit: only 2 of 40 strategies beat buy-and-hold, and a plain sma(50,200) rule outperforms every model's mean Sharpe. Second, self-assessment is badly calibrated: 32 of 40 strategies claim to beat buy-and-hold and exactly one does. Third, swapping the
language-code judge for a second model flips say to do on more than half of items. Injecting Close.shift(-1) drops do to real by 0.33 on average, while our runtime future-information probe fired on neither clean nor injected code. We frame FIDES as a protocol for measurement fidelity, not a claim about
market performance.
Read the full abstract ▾
All authors Arther Tian, Alex Ding, Simon Wu, Aaron Chan
Artificial Intelligence · Cross-listed from cs.HC
Evaluating SAT Solver Metrics as Predictors of Human-Perceived Nonogram Difficulty
Changdao He, Yibing Ju, Jonathan Calver et al. (+1 more)
Algorithmic solver effort is often assumed to align with perceived puzzle difficulty, but this assumption is rarely tested against human solving data. We evaluate this assumption for Nonograms, a popular logic puzzle similar to Sudoku in which numeric clues along each row and column determine a unique solution grid. We formulate Nonograms as a constraint satisfaction problem and solve them using existing SAT solvers. We then conduct a user study in which we collect data on both participant interactions and reported difficulty. We find that neither participants' reported difficulty nor their behavioural signals correlate meaningfully with SAT solver metrics; however, we find evidence that expertise moderates the relationship between solver metrics and reported difficulty. In this process, we uncover distinct, recurring solving strategies that indicate human preference for complex propagation, diverging from solver-measured complexity.
Read the full abstract ▾
All authors Changdao He, Yibing Ju, Jonathan Calver, Alice Gao
Artificial Intelligence · Cross-listed from cs.LG
Sigmoid Attention as a Better Substrate for Learned KV Cache Eviction
Isaac Li
Learned KV-cache eviction often faces a soft-to-hard mismatch: during training, differentiable gates typically attenuate
token contributions, whereas inference saves
memory only when KV entries are physically removed. We ask whether the
attention substrate affects this soft-to-hard transition. Using GPT-2-scale
Transformers trained on OpenWebText, we run a controlled $2\times2\times2$ comparison over
attention type, learned gating, and positional encoding. Although sigmoid
attention is worse as a dense
language model, learned hard eviction changes the useful operating points: sigmoid-gated models delete KV entries with negligible PPL change relative to their own no-eviction references. Under a matched live-cache protocol on the same dense backbones, learned sigmoid gates obtain lower PPL than our H$_2$O and KeyDiff implementations, whereas softmax gates do not uniformly beat these post-hoc methods. The results suggest that
attention normalization can substantially affect whether a training-time soft gate transfers cleanly to hard KV deletion.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.LG
How Much Regularization Survives Averaging? Update Masking in Federated Learning
Wenhao Yan, Fu Kuroda, Yucheng Jin et al. (+1 more)
Federated learning on non-IID data seeks flat minima to generalize across clients, and existing methods borrow sharpness-aware minimization from centralized training. There is a second way to reach flat minima, in which the regularization comes for free from noise added to the parameter updates, and it has never been carried over to the
federated setting as an implicit regularizer. We show the reason. Masking charges the optimizer for moving in sharp directions. We prove that when each client draws its own mask,
federated averaging weakens that charge by exactly the cohort size, and that giving every client the same mask brings it back by a factor equal to the inverse gradient diversity of the cohort. In our experiment setting on CIFAR-10, that factor is 1.19 out of a possible 10. Turning off minibatch sampling raises it to 8.96, while changing data heterogeneity a thousandfold leaves it between 1.17 and 1.50. The configurations keeping the regularization train far too poorly to use.
Read the full abstract ▾
All authors Wenhao Yan, Fu Kuroda, Yucheng Jin, Zhenke Chen
Artificial Intelligence
Apodex 1.1: Scaling Agentic Intelligence for Complex Work
B. An, B. Li, B. Wang et al. (+72 more)
General-purpose
language models can reason and synthesize knowledge, but complex work also requires sustained interaction with files, information sources, and executable code, together with state maintenance, failure recovery, and verifiable delivery. We call this working capability: sustained, verifiable progress toward a real-world objective. Apodex 1.1 develops this capability along two complementary dimensions. Environment Scaling expands the diversity and verifiability of executable file, search, and code environments, while Agentic Coordination Scaling trains
agents to decompose long-horizon tasks, delegate parallel work, integrate asynchronous results, and replan. A shared execution harness and AgentOS maintain task state and provenance across tools and
agents, and training turns environment trajectories and coordination traces into reliable behavior. Across complex professional work, finance, scientific research, mathematics, coding, and search, Apodex 1.1 reaches the leading performance band despite using a substantially smaller model than many frontier systems. The 35B-parameter Apodex 1.1 Mini further retains strong working capability in a locally deployable form. These results ground agentic intelligence in useful, verifiable work completed over time and advance our goal of building a Heavy-Duty Solver for ambitious, long-running tasks.
Read the full abstract ▾
All authors B. An, B. Li, B. Wang, B. Zhang, B.L. Wang, C. Feng, C. Wei, C. Xue, C. Zhang, D. Ng, D. Ye, E. Min, F. Chen, F. Liu, F. Yang, F. Ye, G. Sun, H. Ji, H. Xu, H. Yang, H. Ye, H. Zhang, H. Zhao, J. Li, J. Lin, J. Xia, K. Jin, K. Wang, K. Yang, L. Bing, L. Lei, L. Su, Le. Wang, Lu. Wang, N. Wang, Q. Ren, Q. Yang, R. Li, S. Bai, S. Du, S. Li, S. Lin, S. Nie, S. Wang, S. Zhang, S.Z. Wang, T. Ge, Ta.Q. Fang, Ti.Q. Fang, W. Fang, W. Li, W. Zhang, X. Chen, X. Li, X. Tang, X. Wang, X. Xu, X. Zhang, X.Q. Wang, X.Y. Wang, Y. Deng, Y. Gao, Y. Hu, Y. Li, Y. Sui, Y. Wang, Y. Xiao, Y. Zhang, Y. Zhou, Z. Chen, Z. Cheng, Z. Feng, Z. Liang, Z. Liu, Z. Zhang
Artificial Intelligence
Hidden in the Request: Explaining Unethical LLM Compliance through Token Relevance
Or Biton, Tomer Krichli, Itai Allouche et al. (+1 more)
Although Large
Language Models (LLMs) are aligned to optimize for both helpfulness and harmlessness, these dual objectives may conflict, inevitably leading to
alignment failures. This work systematically investigates instances where LLMs fail to exhibit ethical behavior. To understand the underlying mechanics of these vulnerabilities, we introduce a probing methodology that presents unethical scenarios to LLMs in three distinct structural modalities: objective classification tasks, subjective first-person statements, and direct requests for assistance. We find that model performance degrades in the request-for-assistance-based form. Using Layer-wise Relevance Propagation (LRP), we trace this discrepancy to an attribution bias: the model places greater emphasis on benign task-framing
tokens (e.g., "Can you help me...") than on
tokens signaling the underlying unethical behavior (e.g., "without getting caught"), which we term cue-
tokens. We hypothesize that this under-attribution contributes to harmful compliance. To test this, we introduce two LRP-guided decoding methods that steer generation toward trajectories more relevant to cue
tokens. Empirical evaluations show that these interventions promote safer responses, supporting cue-
token attribution's role in compliance failures.
Read the full abstract ▾
All authors Or Biton, Tomer Krichli, Itai Allouche, Joseph Keshet
Artificial Intelligence
Automated Construction of FAIR Digital Object Knowledge Graphs from Flat Cultural Heritage Records
Zeyd Boukhers, Lingxiao Kong, Xenophon Zabulis et al. (+1 more)
The FAIR Digital Object (FDO) framework mandates that metadata attribute values be expressed as persistent identifiers (PIDs) wherever possible, to produce a fully machine-actionable graph in which every reference is resolvable. The Europeana Data Model was designed long before the FDO specification, and it stores most metadata values as plain text. This serves human browsing well enough, but gives an automated
agent nothing to follow across records or collections. We present a pipeline that transforms flat Europeana records into an FDO-compliant knowledge graph structured with CIDOC-CRM. Following the FDO specification, we model every heritage entity as a discrete FDO with its own PID, type, profile, and metadata layer. The core technical challenge is automating the FDO-prescribed distinction between values that must become PID references (resolvable entities) and those that may remain literals (terminal leaves such as notes, measurements, and dates). We address this with a large
language model that classifies each metadata value, routes it to a controlled vocabulary (Getty AAT, Wikidata, VIAF, PeriodO), and links it to a shared entity FDO. We evaluate using 637 archaeological records from five Europeana providers, processing each with the LLM. The pipeline links 86% of metadata slots, resolving 58.5% of values Europeana had not already enriched. It also merges cross-lingual surface forms that byte-identical matching keeps apart, where 17 of 33 such merges are correct on manual review. Graph connectivity does not separate this from string matching; what distinguishes the FDO graph is that every node is typed and resolvable.
Read the full abstract ▾
All authors Zeyd Boukhers, Lingxiao Kong, Xenophon Zabulis, Georgios Toubekis
Artificial Intelligence · Cross-listed from cs.CV
Progressively Learning Heterogeneous Skills in a Unified Latent Space
Yue-Yi Zhang, Ming Gong, Linpu He et al. (+2 more)
We propose HetSkills, a novel framework designed to progressively learn heterogeneous skills within a unified latent space for physics-based character control. The core idea is to treat this latent space as a shared executable interface, enabling seamless integration of skills learned from diverse data sources, supervision forms, and tasks. HetSkills begins by learning a tracking skill that establishes a strong foundation in motion control and creates a shared motion decoder, which can be reused across tasks without the need for retraining or separate controllers. To prevent the text-to-motion skill from exploiting shortcut pathways instead of learning
language semantics, we introduce motion intuition
distillation to ground text-to-motion generation in
language semantics and a task-guidance module that dynamically adjusts actions based on high-level
language instructions. This enables HetSkills to preserve natural motion while continuously expanding its skill repertoire, making it highly adaptable for long-horizon tasks. Experimental results demonstrate the effectiveness in motion tracking, text-to-motion generation, motion completion, and downstream task adaptation, achieving impressive success rates even under challenging conditions.
Read the full abstract ▾
All authors Yue-Yi Zhang, Ming Gong, Linpu He, Wei-Shi Zheng, Zhilin Zhao
Artificial Intelligence
Is Next-Chunk Reasoning RL Really Better than SFT? Revisiting Training Strategies under no-CoT Data
Yinhao Tang, Youqing Fang, Yanan Sun et al. (+8 more)
Recent work proposes next-chunk
reasoning RL for leveraging no-CoT data—corpora such as worked solutions and textbook derivations that contain
reasoning-rich content but lack explicit chain-of-thought annotations. The method trains a model to generate implicit
reasoning traces and rewards them by their ability to predict the next chunk of text. While promising, existing evaluations primarily compare against conventional SFT baselines, leaving open whether the gains come from the RL formulation itself or from more effectively exposing the model to no-CoT data. We address this question with a controlled study of next-chunk
reasoning RL and a simple but previously overlooked alternative: Mixed SFT, a single supervised fine-tuning stage that jointly trains on no-CoT and long-CoT data. Despite its simplicity, Mixed SFT achieves a clearly higher post-RLVR performance ceiling than next-chunk
reasoning RL while requiring over 60 times less training compute. The advantage is consistent across in-domain mathematical
reasoning and out-of-domain
reasoning tasks. Moreover, we show that higher pre-RLVR accuracy does not necessarily translate into higher post-RLVR accuracy, highlighting the need to evaluate no-CoT training strategies in the context of the full post-training pipeline.
Read the full abstract ▾
All authors Yinhao Tang, Youqing Fang, Yanan Sun, Jiangning Liu, Ziyi Wang, Xun Zhao, Weiming Zhang, Bin Liu, Kuikun Liu, Wenwei Zhang, Kai Chen
Artificial Intelligence · Cross-listed from cs.CV
E2S-Pruner: Progressive Two-Stage Evidence Fusion for Visual Token Pruning in Vision-Language Models
Taoyu Qian, Qi Wang, Daqian Shi et al. (+3 more)
Vision-
language models typically encode an image into hundreds of visual
tokens, incurring substantial inference latency and GPU
memory overhead. Existing
pruning methods largely rely on
attention scores and directly aggregate outputs across
attention heads and network layers, making it difficult to characterize evidential uncertainty and conflict. We propose E2S-Pruner, a progressive two-stage evidence-fusion framework for visual
token pruning that requires no auxiliary model, trainable parameters, or fine-tuning. In the first stage, E2S-Pruner treats each
attention head as an independent evidence source, estimates its reliability from evidence clarity and inter-head consistency, and represents each visual
token using three states: important, unimportant, and uncertain. In the second stage, Dempster–Shafer evidence theory is used to quantify inter-layer conflict and fuse complementary evidence from multiple network layers. We further introduce a spatial novelty constraint that promotes coverage of distinct image regions and prevents the retained
tokens from concentrating in a few locally salient areas. On LLaVA-1.5-7B, E2S-Pruner retains 98.0%, 96.8%, and 90.6% of the aggregate performance when the average numbers of retained visual
tokens are 192, 128, and 64, respectively, while improving throughput by 1.96x and 2.09x under the 128-
token and 64-
token settings. Experiments on Qwen2-VL-7B further demonstrate cross-model generalization. Code is available at https://github.com/taoyu-qian/E2S-Pruner.git.
Read the full abstract ▾
All authors Taoyu Qian, Qi Wang, Daqian Shi, Yuanhao Jiang, Shang Gao, Hualong Yu
Artificial Intelligence · Cross-listed from cs.CL
Future Querying: Can LLMs Serve as Implicit Medical World Models?
Siri Willems, James Butterworth, Lore Goetschalckx et al. (+4 more)
Traditional clinical prediction models rely on task-specific pipelines and curated, structured data, which scale poorly and underutilize unstructured text. To address this, we introduce future querying, a paradigm that probes whether large
language models (LLMs) can function as implicit medical world models by evaluating their ability to answer time-indexed clinical queries about a patient's future. Our framework operates on unstructured clinical documentation using endpoint-agnostic training, enabling a single model to answer diverse clinical queries over patient trajectories without manual feature engineering or task-specific retraining. We show that small, locally fine-tuned open-weight models can match or approach larger proprietary systems, making the framework suitable for
privacy-preserving, on-premise deployment. Evaluated on a new synthetic medical reports
dataset and real ICU notes from the MIMIC-IV
dataset, our results provide encouraging evidence that LLMs can capture aspects of clinical dynamics.
Read the full abstract ▾
All authors Siri Willems, James Butterworth, Lore Goetschalckx, Peter Vrancx, Philippe Modard, Elke Giets, Ludovic Denoyer
Artificial Intelligence · Cross-listed from cs.GT
Multi-Winner Voting with Argumentative Ballots
Ryuta Arisaka and Hirotaka Ono
We introduce multi-winner voting with argumentative ballots (MVArg) and investigate theoretical properties. As our conceptual contribution, we generalise approval ballots to argumentative ballots, thereby allowing voters to express defeasible preferences over candidates. We accordingly generalise voter cohesion and justified representation axioms JR, PJR and EJR. As our theoretical contribution, we establish several key results. First, MVArg is strictly more expressive than multi-winner voting with approval ballots (MV). Second, our notions of cohesion and justified representation are conservative generalisations of their counterparts in MV. Third, the MVArg counterpart of JR can always be satisfied, whereas the counterparts of PJR and EJR cannot always be. Fourth, although verifying whether a winner set satisfies the MVArg counterpart of JR is already coNP-hard, such a winner set can be constructed in
polynomial time. All definitions, propositions, auxiliary lemmas and
theorems have been formalised and mechanically checked in Lean 4.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.CL
Credal Large Language Models for Semantic Commitment under Uncertainty
Shireen Kudukkil Manchingal, Sofiia Nikolenko and Fabio Cuzzolin
Large
language models (LLMs) often produce fluent but incorrect answers with unwarranted confidence. A central limitation is that standard LLMs represent uncertainty through a single predictive distribution, conflating epistemic ignorance with genuine ambiguity. We introduce Credal Large
Language Models (CLLMs): an ensemble of LoRA adapters induces a credal set whose lower and upper probabilities expose the spread of plausible predictive distributions rather than collapsing to a single softmax output. From this representation, we derive a single commitment rule: the model commits to an answer only when its lower probability exceeds the upper probability of every alternative, and otherwise returns the set of answers that no plausible predictor rules out. We apply this commitment rule at two depths: Credal
Token Commitment (CTC) applies it to answer
tokens from one ensemble forward pass, which decides constrained answers without any generation; for open-ended answers, credal decoding extends a partial answer only when no completed answer dominates it, so that the completions produced are those the plausible predictors license, and Credal Semantic Commitment (CSC) applies the rule to their meaning clusters. We evaluate CLLMs with Gemma-2-9B, Llama-3.1-8B and Qwen2.5-7B on OpenBookQA, CoQA, TriviaQA and ARC-Challenge. On multiple choice, CTC commits on 73-91% of questions at 89-98% accuracy, returns sets of 1.1-1.5 options containing the gold one on 89-98%, and its intervals contain the observed accuracy in 24 of 30 confidence bins without calibration; corrupted context lowers commitment from 87-92% to 65-71%, and on Gemma the credal bound detects corruption better than every baseline. On open-ended QA, CLLM outperforms semantic
entropy and Laplace-LoRA at a fixed coverage by up to 19% and 9.5% absolute accuracy on CoQA and TriviaQA with context, for every backbone.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.IR
Retrieval-Augmented Classification of Environmental Mitigations in Hydropower Licensing Documents
Hong-Jun Yoon, Tom Ruggles, Joanna Lee et al. (+1 more)
Identifying and classifying environmental mitigation obligations in Federal Energy Regulatory Commission hydropower licensing documents is a labor-intensive task requiring deep domain expertise. We formulate this as a multi-label classification problem over a structured 135-category taxonomy and address the central challenge of severe label scarcity: 40 of 135 categories have no training examples, and 26 have fewer than five. A supervised Bidirectional Encoder Representations from
Transformers (BERT)-based pipeline, while effective on well-represented categories, achieves F1 of zero on unseen classes regardless of augmentation strategy. We introduce a
Retrieval-Augmented Generation (RAG) pipeline that conditions classification on retrieved category definitions, enabling zero-shot generalization across the full label space. We further propose a hybrid system that combines BERT detection with RAG classification, exploiting the high recall of fine-tuned detection and the zero-shot coverage of
retrieval-augmented
reasoning. Evaluated on the full set of 2017 license documents (5,860 paragraphs, 135 categories), the hybrid achieves a Micro F1 of 0.524, outperforming the BERT-only pipeline (0.477) and the RAG-only pipeline (0.416) across all training-support buckets.
Read the full abstract ▾
All authors Hong-Jun Yoon, Tom Ruggles, Joanna Lee, Debjani Singh