探索全球前沿学术脉络

01.

arXiv (CS.CL) 2026-06-25 DOI: arXiv:2606.25996

Autodata: An agentic data scientist to create high quality synthetic data

作者:

Ilia Kulikov↗Chenxi Whitehouse↗Tianhao Wu↗Yixin Nie↗Swarnadeep Saha↗Eryk Helenowski↗Weizhe Yuan↗Olga Golovneva↗Jack Lanchantin↗Yoram Bachrach↗Jakob Foerster↗Xian Li↗…

We introduce Autodata, a general method that enables AI agents to act as data scientists who build high quality training and evaluation data. We show how to train (meta-optimize) such a data scientist agent, so that it learns to create even stronger data. We describe the overall formulation, and a specific practical implementation, Agentic Self-Instruct. We conduct experiments on computer science research tasks, legal reasoning tasks and reasoning with mathematical objects, where we obtain improved results compared to classical synthetic dataset creation methods. Further, meta-optimizing the data scientist agent itself delivers an even larger performance uplift. Agentic data creation provides a way to convert increased inference compute into higher quality model training. Overall, we believe this direction has the potential to change the way we build AI data.

阅读与讨论 → 访问原文 →

02.

arXiv (CS.CL) 2026-06-12 DOI: arXiv:2606.13317

SkillCAT: Contrastive Assessment and Topology-Aware Skill Self-Evolution for LLM Agents

作者:

Kunfeng Chen↗Qihuang Zhong↗Juhua Liu↗Bo Du↗

Skill self-evolution methods for LLM agents aim to turn execution trajectories into reusable skill documents, but current pipelines typically learn from one trajectory per task, merge candidate skill patches before checking them, and load the full skill corpus before inference. We propose SkillCAT, a training-free framework that separates this process into three stages. Contrastive Causal Extraction (CCE) samples multiple trajectories for each task and compares same-task success/failure pairs to identify evidence that explains outcome differences. Assessment-Augmented Evolution (AAE) replays each candidate patch on source-task clones and keeps only patches that improve or preserve task outcomes before hierarchical skill patch merging. Topology-Aware Task Execution (TTE) compiles the evolved skills into a routable sub-skill topology, so inference loads only the capability nodes relevant to the task. We evaluate SkillCAT on common agent benchmarks, including SpreadsheetBench, WikiTableQuestions, and DocVQA, and further test cross-model and out-of-distribution generalization. Across these settings, SkillCAT raises the average score over baselines by up to 40.40%, demonstrating reliable skill evolution without model training.

阅读与讨论 → 访问原文 →

03.

arXiv (quant-ph) 2026-06-15 DOI: arXiv:2511.19302

Quantitative and Optimal Device-Independent Lower Bounds on Detection Efficiency

作者:

Arkaprabha Ghosal↗Soumyadip Patra↗Peter Bierhorst↗

arXiv:2511.19302v2 Announce Type: replace Abstract: This paper examines a quantitative and optimal lower bound on the detector efficiency in a (2,2,2) Bell experiment within a fully device-independent framework, whereby the detectors used in the experiment are uncharacterized. We provide a tight lower bound on the minimum efficiency required to observe a desired Bell-CHSH violation using the Navascués-Pironio-Acín (NPA) hierarchy, confirming tightness up to four decimal places with numerical optimization over explicit quantum realizations. We then introduce the effect of dark counts and demonstrate how to quantify the minimum required efficiency to observe a desired CHSH violation with an increasing dark count error. Finally, to obtain an analytical closed-form expression of the minimum efficiency, we consider the set of no-signaling behaviors that satisfy the Tsirelson bound, which are easier to characterize than the quantum set. Using such behaviors, we find a simple closed-form expression for a lower bound on the minimum efficiency which is monotonically increasing with the CHSH violation, though the analytically obtained lower bounds are meaningfully below the numerically tight lower bound.

阅读与讨论 → 访问原文 →

04.

medRxiv (Medicine) 2026-06-17 DOI: HASH:b57b4288dd9142d14f75f4a7ba4b834e

Cross-Device Adaptation of Mirai for Mammography-Based Breast Cancer Risk Prediction

作者:

Sistig↗Rothstein↗J. H↗Gadgil↗Achacoso↗Alexeeff↗S. E↗Gerstley↗L. D↗Klein↗R. J↗Margolies↗…

Fine-tuning can adapt pretrained medical imaging models to new clinical datasets, but device-specific domain shifts may limit generalizability. We evaluated Mirai, a mammography-based deep learning model for breast cancer risk prediction, in a large screening cohort containing Hologic and General Electric (GE) full-field digital mammography systems, including GE Premium View (GE PV) and Tissue Equalization (GE TE) post-processing software. Native Mirai showed lower performance on TE images than on Hologic or PV images. Fine-tuning on TE images improved TE performance, particularly for short-term risk prediction, but substantially reduced performance on Hologic images, consistent with catastrophic forgetting. To mitigate this effect, we developed a device-invariant model using interleaved multi-device sampling and conditional adversarial training. This approach largely restored Hologic performance while maintaining improved TE performance, providing better robustness across heterogeneous imaging platforms. Comparison of cumulative and annual risk AUCs over a five-year time horizon further showed that performance gains were driven mainly by short- and intermediate-term predictions. These findings highlight both the value and dangers of device-specific fine-tuning and support balanced domain-adaptation strategies for deploying mammography-based risk models across diverse clinical imaging environments.

阅读与讨论 → 访问原文 →

05.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2505.23666

LoLA: Low-Rank Linear Attention With Sparse Caching

作者:

Luke McDermott↗Robert W. Heath Jr↗Rahul Parhi↗

The per-token cost of transformer inference scales with context length, preventing its application to lifelong in-context learning. Linear attention is an efficient alternative that maintains a constant memory footprint, even on infinite context lengths. While this is a potential candidate for lifelong learning, it falls short in memory capacity. In this paper, we propose LoLA, a training-free augmentation to linear attention that boosts associative recall. LoLA distributes past key-value pairs from context into three memory systems: (i) recent pairs in a local sliding window cache; (ii) difficult-to-memorize pairs in a sparse, global cache; and (iii) generic pairs in the recurrent hidden state of linear attention. We show through ablations that our self-recall error metric is crucial to efficiently manage long-term associative memories. On pass-key retrieval tasks, LoLA improves the base model's performance from 0.6% to 97.4% accuracy. This is achieved with a 4.6x smaller cache than Llama-3.1 8B on 4K context length. LoLA also outperforms other 1B and 8B parameter subquadratic models on zero-shot commonsense reasoning tasks.

阅读与讨论 → 访问原文 →

06.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2509.07605

Beyond Rebalancing: Benchmarking Binary Classifiers Under Class Imbalance Without Rebalancing Techniques

作者:

Ali Nawaz↗Amir Ahmad↗Shehroz S. Khan↗

arXiv:2509.07605v2 Announce Type: replace-cross Abstract: Class imbalance poses a significant challenge to supervised classification, particularly in critical domains like medical diagnostics and anomaly detection where minority class instances are rare. While numerous studies have explored rebalancing techniques to address this issue, less attention has been given to evaluating the performance of binary classifiers under imbalance when no such techniques are applied. Therefore, the goal of this study is to assess the performance of binary classifiers "as-is", without performing any explicit rebalancing. Specifically, we systematically evaluate the robustness of a diverse set of binary classifiers across both real-world and synthetic datasets, under progressively reduced minority class sizes, using one-shot and few-shot scenarios as baselines. Our approach also explores varying data complexities through synthetic decision boundary generation to simulate real-world conditions. In addition to standard classifiers, we include experiments using undersampling, oversampling strategies, and one-class classification (OCC) methods to examine their behavior under severe imbalance. The results confirm that classification becomes more difficult as data complexity increases and the minority class size decreases. While traditional classifiers deteriorate under extreme imbalance, advanced models like TabPFN and boosting-based ensembles retain relatively higher performance and better generalization compared to traditional classifiers. Visual interpretability and evaluation metrics further validate these findings. Our work offers valuable guidance on model selection for imbalanced learning, providing insights into classifier robustness without dependence on explicit rebalancing techniques.

阅读与讨论 → 访问原文 →

07.

arXiv (CS.AI) 2026-06-15 DOI: arXiv:2606.13858

Mood-Aware Music Recommendation: Integrating User Affective Signals into Ranking Systems

作者:

Terence Zeng↗Abhishek K. Umrawal↗

arXiv:2606.13858v1 Announce Type: cross Abstract: Recommendation systems are essential in modern music streaming platforms due to the vast amount of available content. While collaborative filtering is widely used to suggest items based on the preferences of others with similar patterns, it performs poorly in domains where user-item interactions are sparse, such as music. Content-based filtering is an alternative approach that examines the qualities of the items themselves. Genre, instrumentation, and lyrics have been explored; however, relatively little attention has been given to emotion recognition. Since a user's emotional state strongly influences their music choice, incorporating mood signals offers a promising direction for personalization. In this work, we propose a mood-conditioned ranking framework that integrates user affective signals into the recommendation process via softmax-based sampling in the energy-valence space. We evaluate the approach via single-blind experiments in which participants compare recommendations from the proposed system against a baseline. The results indicate improved perceived recommendation quality, providing preliminary evidence for the effectiveness of incorporating mood-based inputs into music recommendations.

阅读与讨论 → 访问原文 →

08.

arXiv (CS.CL) 2026-06-17 DOI: arXiv:2606.17113

The Critical Role of Model Selection in Causal Inference: A Comparative Analysis of Classification Models within the InferBERT Framework for Pharmacovigilance

作者:

Csaba Kiss↗Roland Molontay↗Gabriele Pergola↗

Distinguishing causal adverse drug events (ADEs) from spurious correlations remains a central challenge in pharmacovigilance. The InferBERT framework integrates transformer models with Do-calculus, but its success hinges on the underlying classification model. This study evaluates the impact of model choice in InferBERT, assessing whether simpler models suffice, if domain-specific pre-training helps, whether scaling to LLMs improves causal detection, and the effect of post-hoc calibration. We performed a comparative study on two benchmarks: Analgesics-induced Acute Liver Failure (AILF) and Tramadol-related Mortalities (TRAM). Four models were evaluated-XGBoost (baseline), ALBERT (original InferBERT), BioBERT (biomedical transformer), and Med-LLaMA (medical LLM)-using 5-fold cross-validation repeated over 20 runs. We measured accuracy, Expected Calibration Error (ECE) pre- and post-isotonic regression, and Jaccard concordance of causal terms with PRR, ROR, and EBGM; significance was tested with paired t-tests. BioBERT achieved the highest accuracy on both datasets, while Med-LLaMA underperformed despite its size and parameter-efficient fine-tuning. Domain-specific pre-training was decisive. Calibration improved ECE but had mixed effects on accuracy and causal discovery. BioBERT's superiority also yielded the strongest concordance with traditional pharmacovigilance signals. These results show that domain-specific pre-training provides a clear advantage over simpler baselines and larger LLMs. Investing in manageable, domain-aware models is more effective for computational pharmacovigilance than simply scaling model size.

阅读与讨论 → 访问原文 →

09.

arXiv (CS.AI) 2026-06-15 DOI: arXiv:2606.14498

A Fixed-Point Neural Operator for Size- and Functional-Transferable Hamiltonian Prediction

作者:

Yunhong Lou↗Xihang Yue↗Xinran Wei↗Tianqi Deng↗Linchao Zhu↗

arXiv:2606.14498v1 Announce Type: cross Abstract: Predicting the Kohn-Sham Hamiltonian with machine learning can accelerate density functional theory while retaining access to molecular orbitals, energy levels, and electronic-structure observables that energy-only surrogates cannot resolve. Yet element-wise agreement with the converged Hamiltonian, an implicit fixed point of the self-consistent field iteration, does not determine the occupied subspace that governs orbital energies and densities. Here we present HamEvo, a neural operator that learns the single-step self-consistent update and returns the converged Hamiltonian as its fixed point. HamEvo is pre-trained on intermediate self-consistent trajectories and calibrated at equilibrium with density-matrix supervision. Across benchmarks from MD17 to drug-like QMugs, HamEvo lowers Hamiltonian errors by 35-49% over direct-regression and deep-equilibrium baselines, and predicts QMugs HOMO and LUMO energies with mean absolute errors of 0.036 and 0.053 eV, near the 1 kcal/mol chemical-accuracy scale. Few-shot fine-tuning with only 20 reference conformations extends HamEvo to molecules of up to 122 atoms, well beyond the size range covered by pre-training. With thermal molecular-dynamics sampling, HamEvo captures temperature-dependent HOMO-LUMO gap renormalization beyond the harmonic approximation. Inference is up to 242 times faster than conventional DFT.

阅读与讨论 → 访问原文 →

10.

arXiv (CS.AI) 2026-06-25 DOI: arXiv:2602.05143

CausalRAG2: Hierarchical Causal Knowledge Graph Design for RAG

作者:

Nengbo Wang↗Tuo Liang↗Vikash Singh↗Chaoda Song↗Van Yang↗Yu Yin↗Jing Ma↗Jagdip Singh↗Vipin Chaudhary↗

arXiv:2602.05143v2 Announce Type: replace Abstract: Retrieval augmented generation (RAG) has enhanced large language models by enabling access to external knowledge, with graph-based RAG emerging as a powerful paradigm for structured retrieval and reasoning. However, existing graph-based methods often over-rely on entity-centric node matching and lack explicit causal modeling, leading to unfaithful or spurious answers. Prior attempts to incorporate causality are typically limited to local or single-document contexts and also suffer from information isolation that arises from modular graph structures, which hinders scalability and cross-module causal reasoning. To address these challenges, we propose CausalRAG2, a framework that rethinks knowledge organization for graph-based RAG through causal gating across hierarchical modules. CausalRAG2 explicitly models causal relationships to suppress spurious correlations while enabling scalable reasoning over large-scale knowledge graphs. We also introduce HolisQA, a benchmark for holistic comprehension beyond entity-centric matching. Extensive experiments demonstrate that CausalRAG2 consistently outperforms competitive graph-based RAG baselines across multiple datasets and evaluation metrics. Our work establishes a principled foundation for structured, scalable, and causally grounded RAG systems. Our code and HolisQA benchmark are available at https://github.com/Pwnb/CausalRAG2.

阅读与讨论 → 访问原文 →

11.

medRxiv (Medicine) 2026-06-22 DOI: HASH:8a19fdbecfdfff07ef399b7bc9331d6b

A Randomized, Controlled, Double Blind Clinical Study to Evaluate Use of Hydron Alkaline Ionised Water (HAIW) in Healthy Participants

作者:

Malhotra↗V. K↗Tamoli↗Kalbhor↗Nipanikar↗Dubey↗Singhvi↗Sharma↗

Background and Objectives: Alkaline Ionized Water (AIW) is considered among the highest quality healthy drinking water worldwide and is widely discussed for its various health benefits. Hydron Alkaline Ionized Water (HAIW) is produced through electrolysis, resulting in a stable pH of approximately 9.5 with a negative Oxidation Reduction Potential (ORP), making it an antioxidant beverage. The objective of this study was to evaluate the safety of HAIW and its effects on digestion, sleep, energy, and overall quality of life in healthy participants compared to Packaged Drinking Water (PDW). Materials and Methods: A randomized, controlled, double blind, prospective clinical study was conducted in which a total of 24 healthy participants between the age group of 21 to 40 years were randomized in a 1:1 ratio to either HAIW Group or Packaged Drinking Water Group with equal gender distribution. Participants were hospitalized for 7 days and asked to consume at least 3 litres of the assigned water daily. Primary outcomes were safety-related laboratory parameters and adverse event monitoring. Secondary outcomes included assessment of digestion (appetite, digestion, bowel habits), urine parameters, sleep quality, freshness after waking, fatigue, energy/stamina/strength, quality of life, and global assessment Results: All 24 participants completed the study with no dropouts. Baseline demographics were comparable between the two groups. Assessment of primary safety-related laboratory parameters including Complete Blood count, liver function tests, renal function tests, blood sugar, Electrocardiogram and serum electrolytes showed non-significant change from baseline to 7 days and remained within normal limits in both groups, with non-significant difference between groups (p>0.05). HAIW showed significantly better improvement in appetite, digestion, and bowel habits from Day 2 onwards compared to Packaged drinking water. Sleep quality and freshness after waking up showed significant improvement from Day 3 and Day 2 respectively in the HAIW and PDW group, with significantly better improvement in HAIW group. Fatigue scores showed significant reduction at Day 6 and 7 in both groups with non-significant difference between groups. A total of 5 adverse events were reported (3 in HAIW, 2 in PDW), all unrelated to study products and were mild in nature. Global assessment showed excellent to good overall safety and tolerability in both groups. Conclusion: HAIW was well tolerated by all participants without any adverse effects. All laboratory safety parameters remained within normal range. HAIW demonstrated significant improvements in digestive function (appetite, digestion, bowel habits), sleep quality, and freshness after waking as compared to PDW. The study concludes that HAIW can be safely consumed. HAIW improves digestive and sleep-related functions.

阅读与讨论 → 访问原文 →

12.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2604.00730

A CEFR-Inspired Classification Framework with Fuzzy C-Means To Automate Assessment of Programming Skills in Scratch

作者:

Ricardo Hidalgo-Arag\'on↗Jes\'us M. Gonz\'alez-Barahona↗Gregorio Robles↗

arXiv:2604.00730v2 Announce Type: replace-cross Abstract: Context: Schools, training platforms, and technology firms increasingly need to assess programming proficiency at scale with transparent, reproducible methods that support personalized learning pathways. Objective: This study introduces a pedagogical framework for Scratch project assessment, aligned with the Common European Framework of Reference (CEFR), providing universal competency levels for students and teachers alongside actionable insights for curriculum design. Method: We apply Fuzzy C-Means clustering to 2008246 Scratch projects evaluated via Dr.Scratch, implementing an ordinal criterion to map clusters to CEFR levels (A1-C2), and introducing enhanced classification metrics that identify transitional learners, enable continuous progress tracking, and quantify classification certainty to balance automated feedback with instructor review. Impact: The framework enables diagnosis of systemic curriculum gaps-notably a "B2 bottleneck" where only 13.3% of learners reside due to the cognitive load of integrating Logic Synchronization, and Data Representation–while providing certainty–based triggers for human intervention.

阅读与讨论 → 访问原文 →

13.

arXiv (CS.AI) 2026-06-25 DOI: arXiv:2606.20615

Specifying AI-SDLC Processes: A Protocol Language for Human-Agent Boundaries

作者:

Ylli Prifti↗

arXiv:2606.20615v2 Announce Type: replace Abstract: AI agents now participate as first-class team members across the software development lifecycle, yet no specification language exists for expressing the human-agent responsibility boundaries, approval gates, and governance constraints this collaboration requires. Existing approaches encode process in agent prompts (subject to drift), target adjacent domains (workflow management, business processes), or address only fragments (access control, approval gates). We propose a domain-specific language for specifying AI-SDLC processes as protocols, with formal syntax, well-formedness conditions, operational semantics, and enforcement invariants. The language distinguishes policy (declared intent) from mechanism (structural enforcement), enabling implementations to bound process non-determinism through primitives such as validation tokens and capability boundaries. Three results follow. A failure rate analysis shows that structural enforcement bounds system failure rates at a weighted product of agent and validator rates, while behavioral compliance permits cumulative or near-saturating growth. The 2+N team pattern (two human-in-control roles plus N specialized agent members) formalizes classical Separation of Duties for AI-SDLC. Kleene closure of orchestration loops and reflexive protocol-adherence validation emerge as design properties rather than special-case constructs. We position the contribution against multi-agent frameworks (MetaGPT), workflow specification (FlowAgent, BPMN extensions), and capability-based security (SAGA): the novelty lies in the specific integration, not any single primitive. A working implementation demonstrates feasibility; empirical evaluation is future work.

阅读与讨论 → 访问原文 →

14.

arXiv (CS.CV) 2026-06-25 DOI: arXiv:2603.14609

GroundSet: A Cadastral-Grounded Dataset for Spatial Understanding with Vector Data

作者:

Roger Ferrod↗Ma\"el Lecene↗Krishna Sapkota↗George Leifman↗Vered Silverman↗Genady Beryozkin↗Sylvain Lobry↗

Precise spatial understanding in Earth Observation is essential for translating raw aerial imagery into actionable insights for critical applications like urban planning, environmental monitoring and disaster management. However, Multimodal Large Language Models exhibit critical deficiencies in fine-grained spatial understanding within Remote Sensing, primarily due to a reliance on limited or repurposed legacy datasets. To bridge this gap, we introduce a large-scale dataset grounded in verifiable cadastral vector data, comprising 3.8 million annotated objects across 510k high-resolution images with 135 granular semantic categories. We validate this resource through a comprehensive instruction-tuning benchmark spanning seven spatial reasoning tasks. Our evaluation establishes a robust baseline using a standard LLaVA architecture. We show that while current RS-specialized and commercial models (e.g., Gemini) struggle in zero-shot settings, high-fidelity supervision effectively bridges this gap, enabling standard architectures to master fine-grained spatial grounding without complex architectural modifications.

阅读与讨论 → 访问原文 →

15.

arXiv (CS.CL) 2026-06-11 DOI: arXiv:2606.11709

RLCSD: Reinforcement Learning with Contrastive On-Policy Self-Distillation

作者:

Leyi Pan↗Shuchang Tao↗Yunpeng Zhai↗Lingzhe Zhang↗Zhaoyang Liu↗Bolin Ding↗Aiwei Liu↗Lijie Wen↗

On-policy self-distillation (OPSD) provides dense, token-level supervision for reasoning models by aligning a model's own distribution with the distribution it produces under privileged context, typically a verified solution. However, we show that the learning signal drawn from this distributional gap concentrates on style tokens rather than task-bearing ones, as the hinted model tends to produce more direct, shorter outputs. We term this pathology privilege-induced style drift, which destabilizes training or causes response length to shrink. To address this, we propose RLCSD (Reinforcement Learning with Contrastive on-policy Self-Distillation), which mitigates this drift by contrasting the teacher-student gap under a correct hint against that under a wrong hint, suppressing the style shift that conditioning on a hint tends to induce regardless of correctness, and yielding a signal that is more concentrated on task-bearing tokens. Experiments on Qwen3 (1.7B/4B/8B) and Olmo-3-7B-Think across mathematical and logical reasoning show that RLCSD consistently outperforms GRPO and prior OPSD methods. We further show that the contrastive principle is general: it plugs into existing OPSD methods to improve them, and its underlying insight extends to the broader cross-model on-policy distillation setting.

阅读与讨论 → 访问原文 →

16.

medRxiv (Medicine) 2026-06-24 DOI: HASH:b52131b0c56c7f0d956b2e6d24a846d6

Screen-Free Haptic Breathwork with HRV-Adaptive Control, Pilot Outcomes and System Design

作者:

Adhia↗Raithatha↗Ferguson↗Pasquier↗

Vayu is a mobile breathwork system comprising an iOS companion app and Apple Watch application that delivers slow, resonant breathing using screen-free haptic cues, HRV-adaptive pacing, and reflective journaling grounded in Patanjali's five states of mind. The watchOS component provides tactile phase guidance and real-time biometric sensing (heart rate, HRV), while the iOS interface supports analytics and personalized recommendations. In a 4-6-week naturalistic pilot involving 199 adults (ages 22-65) across Canada, the United States, and India, participants engaged in daily 5-10-minute sessions guided by on-wrist haptics. Average adherence was 4.1 +/- 2.3 sessions per week, with 71% of active users maintaining at least 3 sessions per week. By week four, perceived stress (PSS-10) decreased by 2.5 points, resting heart rate declined by 7.4 bpm, and HRV increased by a median of 28.6% relative to baseline, accompanied by mood improvements. No adverse events were reported. HRV metrics are derived from Apple Watch PPG-based proxies and interpreted as relative trends. These findings suggest Vayu is effective and well-tolerated, demonstrating strong engagement and early efficacy signals.

阅读与讨论 → 访问原文 →

17.

arXiv (CS.AI) 2026-06-25 DOI: arXiv:2601.16529

SycoEval-EM: Sycophancy Evaluation of Large Language Models in Simulated Clinical Encounters for Emergency Care

作者:

Dongshen Peng↗Yi Wang↗Austin Schoeffler↗Sun-ha Hong↗Brian Suffoletto↗David Kim↗Carl Preiksaitis↗Christian Rose↗

arXiv:2601.16529v3 Announce Type: replace Abstract: Large language models (LLMs) deployed in clinical decision support may acquiesce to patient requests for care that conflicts with evidence-based guidelines. We developed SycoEval-EM, a multi-agent simulation framework to evaluate LLM robustness to adversarial patient persuasion in emergency medicine. Across 19 contemporary LLMs and 1,425 simulated clinical encounters spanning three Choosing Wisely scenarios, acquiescence rates ranged from 0% to 100%, revealing a bimodal distribution. Seven models maintained near-perfect guideline adherence, while six acquiesced in the majority of encounters. Vulnerability varied substantially across clinical scenarios. Acquiescence was highest for CT imaging requests, intermediate for antibiotic prescriptions for sinusitis, and lowest for opioid prescriptions for acute back pain. Model scale, recency, and performance on static medical benchmarks did not consistently predict robustness. All five persuasion tactics produced similar acquiescence rates, with no statistically significant differences after correction for multiple comparisons, suggesting a generalized susceptibility rather than tactic-specific weaknesses. LLM-as-judge evaluation was validated against two independent physician raters across 95 matched conversations and demonstrated near-perfect agreement for the primary outcome of acquiescence (Cohens kappa = 0.957). These findings indicate that static medical benchmarks are insufficient to predict safety performance under sustained social pressure and support incorporating multi-turn adversarial testing into clinical AI evaluation. Notably, two models achieved perfect guideline adherence across all encounters, demonstrating that robustness to patient pressure is attainable without sacrificing effective clinical communication.

阅读与讨论 → 访问原文 →

18.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2606.18988

ThinkDeception: A Progressive Reinforcement Learning Framework for Interpretable Multimodal Deception Detection

作者:

Jinhao Song↗Shan Liang↗Yiqun Yue↗Zhuhuayang Zhang↗Tianqi Gao↗

arXiv:2606.18988v1 Announce Type: new Abstract: Multimodal deception detection is critical for identifying fraudulent intentions, yet existing approaches predominantly rely on end to end black–box paradigms. These methods suffer from a severe lack of interpretability failing to provide transparent reasoning trajectories and struggling to explicitly capture the subtle, cross modal inconsistencies inherent in deceptive behaviors. To transcend these limitations, we propose ThinkDeception, a novel and interpretable multimodal deception detection framework. As a pioneering effort, it introduces Multimodal Large Language Models (MLLMs) into this domain, transforming deception detection from a traditional binary classification task into an explicit cognitive reasoning process. Facilitated by the first meticulously annotated step–by–step multimodal Chain of Thought (CoT) dataset, we develop a foundational model, ThinkDeception Base, empirically validating the critical role of modal inconsistency in decoding deception. Building upon this foundation, our core innovation lies in proposing Visual-Audio Consistency Group Relative Policy Optimization(VAC–GRPO) equipped with a progressive training strategy. Distinct from standard GRPO, we stratify the training data into four progressive difficulty tiers, guiding the model through a psychologically grounded easy–to–hard cognitive transition. By innovatively coupling this dynamic curriculum scheduler with a multi dimensional, process aware reward mechanism and a reflective learning paradigm, we significantly elevate the model's overall reasoning quality. Extensive experiments on mainstream benchmarks demonstrate that ThinkDeception establishes a new SOTA, significantly outperforming existing methods in both detection accuracy and rationale quality. Ultimately, this work successfully drives the field of deception detection toward interpretable, multimodal cognitive reasoning.

阅读与讨论 → 访问原文 →

19.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.15912

On-Policy Distillation with Curriculum Turn-level Guidance for Multi-turn Agents

作者:

Gengsheng Li↗Mao Zheng↗Mingyang Song↗Ruiqi Liu↗Tianyu Yang↗Jie Sun↗Qiyong Zhong↗Haiyun Guo↗Junfeng Fang↗Dan Zhang↗Jinqiao Wang↗

arXiv:2606.15912v1 Announce Type: cross Abstract: Multi-turn agents that plan, invoke tools, and interact with environments offer a promising paradigm for solving complex tasks, yet their capabilities typically rely on very large models whose inference cost is prohibitive in practice.On-Policy Distillation (OPD) is a natural recipe for transferring such capabilities to smaller students, but we find that it suffers a characteristic failure mode in this setting: small student errors compound across turns and push the trajectory out of the teacher's familiar state distribution, so the teacher's supervision becomes least reliable precisely where the student needs it most.We propose Guided On-Policy Distillation (Guided-OPD), a simple yet effective algorithm that mixes teacher- and student-generated turns within each rollout and schedules the teacher's intervention probability along a curriculum that decays to zero.Strong guidance keeps early trajectories close to the teacher distribution and is then gradually withdrawn to recover the purely on-policy regime used at inference.On ALFWorld, ScienceWorld, and WebShop, distilling Qwen3 students from a Qwen3-30B-A3B teacher, Guided-OPD improves Score by 21.1\% and Success Rate by 25.5\% over vanilla OPD on average, with larger gains on smaller students.

阅读与讨论 → 访问原文 →

20.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.16364

Looking Is Not Picking: An Attention-Segment Account of Tool-Selection Failures in LLM Agents

作者:

Shiyang Chen↗

arXiv:2606.16364v1 Announce Type: new Abstract: LLM agents mis-call tools, and the natural guess is that the model failed to see the right tool in a crowded harness. We show the opposite through a lens concurrent work sets aside – the model's attention to labeled tool-definition segments. On real BFCL failures, by per-candidate attention argmax the model attends most to the correct tool 80% of the time (vs. 21% chance), and the gold is the under-attended segment on only 10%: it looks at the right tool and still picks wrong. This directly refutes the intuitive "crowded-harness / lost-in-the-middle" explanation: the failure is at the decision readout, not the harness, and we pin it there three ways. (1) Input vs. readout: repairing the prompt (reordering or duplicating the gold tool) recovers

阅读与讨论 → 访问原文 →

21.

arXiv (CS.CL) 2026-06-18 DOI: arXiv:2606.18699

TW-LegalBench: Measuring Taiwanese Legal Understanding

作者:

Fei-Yueh Chen↗Chun Huang Lin↗Chan Wei Hsu↗Kuan Hsuan Yeh↗Zih-Ching Chen↗Kuan-Ming Chen↗Patrick Chung-Chia Huang↗

Large language models (LLMs) have shown impressive capabilities across diverse tasks, yet their performance on jurisdiction-specific legal reasoning remains underexplored. We present TW-LegalBench that utilizes Taiwanese legal system's rich official corpus open to the public to fill the gap in evaluating LLMs on Taiwanese law, among common-law benchmarks that focus on English sources and civil-law benchmarks focusing on sources of Simplified Chinese. TW-LegalBench comprises three task types: (1) over 16,000 multiple-choice questions (MCQs) across five years of official examinations in 18 professional domains; (2) 117 open-ended essay questions (OEQs) from examinations for legal professionals with official scoring rubrics; and (3) more than 14,000 legal judgment prediction (LJP) instances covering hundreds of crime categories. We evaluate 13 LLMs using accuracy for MCQs, a decomposed LLM-as-Judge framework based on the scoring rubric points for OEQs, and metrics for sentencing accuracy and statute citation for LJP. Our results reveal that top-performing models exceed the passing threshold for qualified lawyers (passing rate: 11%) but fall short of that for judges and prosecutors (passing rate: 1~2%). For LJP, while models demonstrate reasonable verdict type accuracy and sentence prediction capability, they struggle to cite exact legal articles. These findings highlight that reliable legal text generation remains challenging for LLMs, even though their performance on qualification examinations approaches human level.

阅读与讨论 → 访问原文 →

22.

arXiv (CS.CV) 2026-06-16 DOI: arXiv:2606.16470

Decoupled Object-Centric Video Understanding for Generating Robotic Manipulation Commands

作者:

Thanh Nguyen Canh↗Thanh-Tuan Tran↗Haolan Zhang↗Ziyan Gao↗Xiem HoangVan↗Nak Young Chong↗

Translating video demonstrations into executable robot commands remains challenging because existing methods often fail to identify which objects are functionally involved in the demonstrated action. As a result, they may generate commands that are linguistically plausible but operationally ambiguous. We propose an object-centric video understanding framework that decouples action recognition from object identification to generate precise, grammar-free manipulation commands. Our approach integrates Temporal Shift Modules (TSM) for efficient spatio-temporal action classification with a novel Object Selection algorithm that identifies task-relevant objects through trajectory-based role classification, blur detection, and overlap minimization. The selected objects are then processed by Vision-Language Models (VLMs) for robust category recognition and zero-shot generalization. Evaluated on a modified Something-Something V2 dataset, our method achieves 86.79\% action classification accuracy and BLEU-4 scores of 0.337 on standard objects and 0.261 on novel objects. These results improve over the strongest task-specific baseline by 80.2\% and 143.9\%, respectively. Larger gains are observed in METEOR and CIDEr, reaching 157.9\% and 171.7\% on novel objects. Across all semantic metrics, our approach consistently outperforms task-specific methods and remains competitive with, or surpasses, large general-purpose VLMs while retaining a modular, object-centric design.

阅读与讨论 → 访问原文 →

23.

arXiv (CS.CL) 2026-06-15 DOI: arXiv:2606.03085

Multi-component Causal Tracing in Large Language Models

作者:

Zirui Yan↗Dennis Wei↗Dmitriy A. Katz↗Prasanna Sattigeri↗Ali Tajer↗

Causal tracing systematically intervenes on a large language model's (LLM's) internal representations to uncover and quantify the causal pathways linking specific inputs or computations to specific metrics of interest, quantifying the LLM's behavior. Building on previous single-component or single-layer studies, this paper presents a unified framework for causally tracing multiple components simultaneously. This framework systematically identifies the subsets of components (e.g., attention heads and multi-layer perceptron neurons) most critical to a desired target performance metric (e.g., accuracy and fairness). This is achieved by incorporating flexible interventions applied to a wide range of desired metrics. To address the combinatorial complexity of the multi-component problem, an efficient algorithm is designed that leverages soft interventions and a carefully designed metric transformation, converting the combinatorial search problem into a continuous one that can be solved efficiently under proper constraints, thereby generating proper binary decisions for selecting components. Experimental results demonstrate that the proposed method efficiently identifies subsets of the model's components that have a high impact on the target metric, outperforming existing baseline approaches. Our code is available at https://github.com/ZiruiYan/multi-component-causal-tracing.

阅读与讨论 → 访问原文 →

24.

arXiv (CS.CV) 2026-06-16 DOI: arXiv:2602.15720

ToaSt: Token Channel Selection and Structured Pruning for Efficient ViT

作者:

Hyunchan Moon↗Cheonjun Park↗Steven L. Waslander↗

Vision Transformers (ViTs) have achieved remarkable success across various vision tasks, yet their deployment is often hindered by prohibitive computational costs. While structured weight pruning and token compression have emerged as promising solutions, they suffer from prolonged retraining and inter-layer dependencies that complicate optimization, respectively. We propose ToaSt, a decoupled framework applying specialized strategies to distinct ViT components. We apply coupled head-wise structured pruning to Multi-Head Self-Attention modules, leveraging attention operation characteristics to enhance robustness. For Feed-Forward Networks (over 60% of FLOPs), we introduce Token Channel Selection (TCS), a training-free method that filters redundant noise channels at inference time. Extensive evaluations across nine diverse models, including DeiT, ViT-MAE, and Swin Transformer, demonstrate that ToaSt achieves superior trade-offs between accuracy and efficiency, consistently outperforming existing baselines. On ViT-MAE-Huge, ToaSt achieves 88.52% accuracy (+1.64%p) with 39.4% FLOPs reduction. ToaSt also transfers effectively to diverse downstream tasks (COCO detection, ADE20K segmentation, CIFAR-100 classification), achieving 52.2 versus 51.9 mAP on COCO. Code: github.com/SHANNonLab-HUFS/ToaSt

阅读与讨论 → 访问原文 →

25.

arXiv (CS.CV) 2026-06-17 DOI: arXiv:2603.02803

Structure-Aware Text Recognition for Ancient Greek Critical Editions

作者:

Nicolas Angleraud↗Antonia Karamolegkou↗Beno\^it Sagot↗Thibault Cl\'erice↗

Recent advances in visual language models (VLMs) have transformed end-to-end document understanding. However, their ability to interpret the complex layout semantics of historical scholarly texts remains limited. This paper investigates structure-aware text recognition for Ancient Greek critical editions, which have dense reference hierarchies and extensive marginal annotations. We introduce two novel resources: (i) a large-scale synthetic corpus of 185,000 page images generated from TEI/XML sources with controlled typographic and layout variation, and (ii) a curated benchmark of real scanned editions spanning more than a century of editorial and typographic practices. Using these datasets, we evaluate three state-of-the-art VLMs under both zero-shot and fine-tuning regimes. Our experiments reveal substantial limitations in current VLM architectures when confronted with highly structured historical documents. In zero-shot settings, most models significantly underperform compared to established off-the-shelf software. Nevertheless, the Qwen3VL-8B model achieves state-of-the-art performance, reaching a median Character Error Rate of 1.0\% on real scans. These results highlight both the current shortcomings and the future potential of VLMs for structure-aware recognition of complex scholarly documents.

阅读与讨论 → 访问原文 →