探索全球前沿学术脉络

01.

medRxiv (Medicine) 2026-06-12 DOI: HASH:899a0673fe5b3245d66ca5ccefd57248

Home-based binocular serious games in virtual reality to treat visual acuity and stereovision in residual amblyopia: AMBER study

作者:

Aurilia↗Martin↗N.-L↗Simon-Martinez↗Antoniou↗M.-P↗Bouthour↗Bavelier↗Backus↗B. T↗Dornbos↗Blaha↗…

Objectives: Amblyopia is a pediatric visual disorder traditionally treated by patching the fellow eye, though many patients retain residual amblyopia post-treatment. Increasing evidence suggests that visual plasticity allows treat-ment beyond the classical therapeutic window. AMBER evaluated the efficacy of binocular serious games in virtual reality (VR) in residual amblyopia. Methods and Analysis: The monocentric, prospective, randomized, crossover trial (reported as case series) includ-ed 14 anisometropic, strabismic, or mixed residual amblyopia patients (6-35 years; 5 children, 9 adults). Participants underwent two 2-month intervention phases: optical correction (standard care) and standard care plus VR games (2.5 h/week), each with a 2-month follow-up. Best-corrected visual acuity (BCVA), stereoacuity, and reading speed were assessed (5 timepoints) using the Sloan and Landolt charts, the Titmus, TNO, Lang II, Asteroid, and Mnread tests. Compliance and adverse events (AE) were recorded. Results: VR training improved BCVA in 10 amblyopic eyes (Landolt and Sloan), with more pronounced effects in anisometropic patients. Six patients showed improved stereoacuity (Titmus; 4x mixed, 1x anisometropic, 1x stra-bismic amblyopia), persistent only in children (1x strabismic, 1x mixed amblyopia). Four improvements were ob-served with TNO (1x), Lang II (1x), Asteroid (0x), and MNread (1x). Despite positive trends, when comparing re-sults of individual patients, between both eyes, and with standard treatment, consistency of improvements cannot be conclusively demonstrated. One non-severe AE (dizziness) was reported. Conclusions: Following individual cases, VR training improved BCVA and stereoacuity, particularly in children and patients with high compliance. However, considering the cohort as a whole, consistency of effects has to be confirmed in larger groups. Thus, the methodologically sophisticated AMBER study revealed differences in VR treatment efficacy between amblyopia types, children/adults, endpoints and tests, offering precious data for the design of meaningful future studies. It shows that neurovisual plasticity gauged by VR-games offers safe, engaging treatment options for residual amblyopia.

阅读与讨论 → 访问原文 →

02.

arXiv (CS.LG) 2026-06-18 DOI: arXiv:2606.18539

TS-Fault: Benchmarking Time Series Forecasters Against Structural Faults

作者:

Yuyang Zhao↗Lian Xu↗Hao Miao↗Chenxi Liu↗Hao Xue↗

arXiv:2606.18539v1 Announce Type: new Abstract: Time series forecasting (TSF) underpins consequential decisions in energy, transportation, finance, and healthcare, yet TSF models are almost universally ranked by a single number (e.g., average error) on clean held-out data, under the implicit assumption that it predicts deployed reliability. However, real faults are not i.i.d noise but structured events with temporal shape, broken cross-variable dependencies, regime change coupled with missingness, and causal propagation across a sensing pipeline. Treating TSF robustness as a data-quality problem, we present TS-Fault, a benchmark that evaluates forecasting models under explicit, parameterized fault scenarios with controllable semantic difficulty. TS-Fault organizes recurring failures into four modes along two orthogonal axes (observation- vs mechanism-level; univariate vs multivariate) and injects each fault into the most prediction-critical window via a unified importance score. This design enables robustness to be tested against the structures models actually rely on, rather than reduced to generic noise sensitivity. We evaluate 21 models across 6 datasets, 4 modes, and 5 difficulty levels under a paired clean/corrupt protocol. The results reveal three findings that contradict common leaderboard intuition: (i) clean-data accuracy anti-correlates with robustness; (ii) clean rankings are preserved under observation-level faults but reshuffled under mechanism-level faults; and (iii) all catastrophic failures occur under mechanism-level faults, with foundation models achieving the highest clean-data accuracy yet exhibiting the greatest fragility. The code is publicly available at https://github.com/Ray-zyy/TS-Fault.

阅读与讨论 → 访问原文 →

03.

arXiv (CS.LG) 2026-06-19 DOI: arXiv:2606.20008

VIMPO: Value-Implicit Policy Optimization for LLMs

作者:

Zhewei Kang↗Aosong Feng↗Sergey Levine↗Dawn Song↗Xuandong Zhao↗

arXiv:2606.20008v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards has become a central tool for improving the reasoning ability of large language models, but current methods face a trade-off between simplicity and credit assignment. Group-relative methods such as GRPO avoid training a critic, but typically assign a trajectory-level advantage to every token. Actor-critic methods provide denser learning signals, but require a learned value function with its own training instability. We introduce VIMPO, a critic-free policy optimization method that derives a policy-implied value function from the optimality conditions of KL-regularized reinforcement learning. For autoregressive generation, the resulting value recurrence can be written in terms of policy-reference log-ratios and anchored by the terminal condition that no future reward remains at the end of a trajectory. This gives a simple value loss that incorporates outcome-level verifiable rewards without training a critic. The same derivation also yields a critic-free actor advantage, allowing VIMPO to separate reward incorporation through the value loss from policy improvement through a PPO-style actor update. On mathematical RLVR benchmarks, VIMPO improves over GRPO across MATH-500, AIME 2024, AIME 2025, and OlympiadBench, with especially larger gains on competition-style evaluations. Under noisy rewards, VIMPO retains a consistent advantage over GRPO, suggesting that policy-implied value optimization can provide finer credit assignment while preserving the practical simplicity of critic-free training.

阅读与讨论 → 访问原文 →

04.

arXiv (CS.CL) 2026-06-19 DOI: arXiv:2606.20002

Connect the Dots: Training LLMs for Long-Lifecycle Agents with Cross-Domain Generalization Via Reinforcement Learning

作者:

Yanxi Chen↗Weijie Shi↗Yuexiang Xie↗Boyi Hu↗Yaliang Li↗Bolin Ding↗Jingren Zhou↗

This work presents a general framework for training large language models (LLMs) to "Connect the Dots" (CoD), a meta-capability required by long-lifecycle agents: as an LLM-based AI agent gets deployed in an environment, it solves a long sequence of tasks while continuously exploring the environment, learning from its own experiences, and iteratively self-updating its context about the environment, thereby achieving progressively better performance on future tasks conditioned on the updated context. Major components of the CoD framework include: (1) algorithm design and infrastructure for end-to-end reinforcement learning (RL) with long rollout sequences interleaving solve-task and update-context episodes; (2) tasks and environments for incentivizing and eliciting the targeted meta-capability in LLMs during training, as well as for faithfully measuring progress during evaluation. We present proof-of-concept implementations of the CoD framework, including a GRPO-style RL algorithm with fine-grained credit assignment, as well as tasks and environments tailored to the targeted meta-capability (rather than domain-specific LLM capabilities or standard task-by-task RL). Empirical results validate the efficacy of end-to-end RL training in the CoD setting, and demonstrate the potential for out-of-distribution generalization – within the training domains, across different domains, and from CoD to Ralph-loop settings – of the elicited meta-capability. Our investigation of CoD connects several lines of prior works, and opens up new opportunities for advancing LLMs and AI agents. To facilitate further research and applications, we release our implementations at \url{https://github.com/agentscope-ai/Trinity-RFT/tree/research/cod/examples/research_cod}.

阅读与讨论 → 访问原文 →

05.

arXiv (CS.CL) 2026-06-18 DOI: arXiv:2606.18910

REVES: REvision and VErification–Augmented Training for Test-Time Scaling

作者:

Yuanxin Liu↗Ruida Zhou↗Xinyan Zhao↗Amr Sharaf↗Hongzhou Lin↗Arijit Biswas↗Mohammad Ghavamzadeh↗Zhaoran Wang↗Mingyi Hong↗

Test-time scaling via sequential revision has emerged as a powerful paradigm for enhancing Large Language Model (LLM) reasoning. However, standard post-training methods primarily optimize single-shot objectives, creating a fundamental misalignment with multi-step inference dynamics. While recent work treats this as multi-turn reinforcement learning (RL), conventional approaches optimize over the multi-step trajectories directly, failing to further exploit the high-quality mistakes in intermediate steps that model can learn from correcting them. We propose a two-stage iterative framework that alternates between online data/prompt augmentation and policy optimization. By converting the intermediate steps (``near-miss'' answers) in the successful recovery trajectories into decoupled revision and verification prompts, our approach concentrates training on both effective answer transformation and error identification. This approach enables efficient off-policy data generation and reduces the computational overhead of long-horizon sampling compared to standard multi-turn RL. On LiveCodeBench, using publicly available test cases as feedback, we observe gains of +6.5 points over the RL baseline and +4.0 points over standard multi-turn training. Beyond coding, our approach matches the previously reported SOTA result on circle packing while using the smallest base model (4B) and far fewer rollouts than the much larger evolutionary search systems. Math results under ground-truth verification further confirm improved correction ability. It also generalizes to out-of-distribution constraint-satisfaction puzzles such as n\_queens and mini\_sudoku, where correctness is defined entirely by problem constraints. Code is available at https://github.com/yxliu02/REVES.git.

阅读与讨论 → 访问原文 →

06.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2606.16591

SING: Synthetic Intention Graph for Scalable Active Tool Discovery in LLM Agents

作者:

Qiao Xiao↗Haochen Shi↗Yisen Gao↗Wenbin Hu↗Huihao Jing↗Tianshi Zheng↗Baixuan Xu↗Ziheng Zhang↗Weiqi Wang↗Haoran Li↗Jiaxin Bai↗Yangqiu Song↗…

Large language model (LLM) agents increasingly rely on agent harnesses that manage context, tools, and multi-turn execution, making tools a central interface for acting in realistic digital environments. As harness-connected tool ecosystems expand to hundreds or thousands of APIs, services, and task-specific skills, exhaustive tool schema injection becomes costly and imposes a closed-world assumption that limits agents to a predefined static inventory. Retrieval-augmented tool selection offers a natural alternative, but existing one-shot retrieval methods often fail to align isolated tool descriptions with the agent's true task intention, especially in long-horizon tasks where required capabilities emerge through decomposition, observations, and newly induced subgoals. We propose SING, an intention-aware active tool discovery framework that builds an intention-tool graph linking user intentions, tool capabilities, and tool collaboration patterns, and dynamically retrieves tools according to evolving task states. Using a unified corpus of 7,471 tools, we evaluate SING on three real-world tool-use benchmarks. SING improves Global Recall@5 by up to 59.8% and downstream success rate by up to 28.9% over baselines, while reducing full-corpus tool-schema exposure by 99.8%, demonstrating that intention-aware graph structure enables more accurate and context-efficient tool discovery in large-scale agentic ecosystems.

阅读与讨论 → 访问原文 →

07.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2606.18936

SciRisk-Bench: A Risk-Dimension-Aware Benchmark for AI4Science Safety

作者:

Linghao Feng↗Yinqian Sun↗Dongqi Liang↗Sicheng Shen↗Chenfei Yan↗Yuxuan Peng↗Yilin Zhao↗Haibo Tong↗Kai Li↗FeiFei Zhao↗Yi Zeng↗

arXiv:2606.18936v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly embedded in AI for Science (AI4Science) workflows, from scientific question answering and literature analysis to laboratory planning and autonomous discovery. This progress creates an urgent need for safety benchmarks that evaluate not only scientific competence, but also whether models recognize and avoid risks in high-stakes scientific contexts. Existing AI4Science safety datasets cover several disciplines and task formats, leaving the underlying risk dimensions underspecified. We introduce SciRisk-Bench, a benchmark designed to evaluate AI4Science safety from two complementary perspectives: explicit risk dimensions and scientific disciplines. SciRisk-Bench covers 7 disciplines, 31 subdisciplines and 10 risk dimensions. In the experimental section, we evaluate both mainstream LLMs and science-oriented LLMs across risk dimensions, disciplines, and sub-disciplines, enabling fine-grained diagnosis of where scientific models remain unsafe.

阅读与讨论 → 访问原文 →

08.

arXiv (CS.LG) 2026-06-16 DOI: arXiv:2606.16900

Factorized Neural Operators Decompose Dynamic and Persistent Responses

作者:

Hao Tang↗Yuechen Duan↗Jiongyu Zhu↗Zimeng Feng↗Hao Li↗Chao Li↗

arXiv:2606.16900v1 Announce Type: new Abstract: Physical systems often exhibit heterogeneous mechanisms, where rapidly evolving dynamics coexist with persistent structures. Capturing such multiscale physical behavior remains challenging for existing neural operators, which typically rely on single dominant inductive bias and therefore couple distinct physical responses into a shared representation. We introduce the Unified Green's Function Framework across domains and propose the Factorized Neural Operators (FaNO), which decompose spectral representations into equivariant dynamic responses and invariant persistent responses, leading to better interpretability and generalization. Mechanistically, we show that the two operator branches spontaneously specialize into distinct physical roles that remain consistent across scales and domains: the equivariant branch captures rapidly varying transient dynamics, whereas the invariant branch extracts coherent persistent structures. This factorized mechanism of FaNO improves prediction accuracy, parameter efficiency and cross-scale generalization across physical systems and domains. In particular, it maintains consistent predictions under long-horizon autoregressive rollout, cross-resolution extrapolation and physical-regime shifts. These findings suggest that scalable physical modeling may benefit from moving beyond single-inductive-bias formulations toward factorized operator representations that better reflect the heterogeneous organization of physical systems, accelerating the reliable deployment of machine learning for scientific computing and discovery.

阅读与讨论 → 访问原文 →

09.

arXiv (CS.CL) 2026-06-18 DOI: arXiv:2606.18530

Evaluating Prompting-Based Defenses Against Domain-Camouflaged Injection Attacks

作者:

Aaditya Pai↗

Domain-camouflaged injection attacks embed malicious instructions in retrieved content using domain-appropriate vocabulary, evading standard detectors that rely on syntactic injection markers. When detection fails, practitioners need to know which defense architectures reduce attack success. We evaluate five prompting-based defenses (spotlighting, paraphrasing, prompt sandwiching, and two combinations) against domain-camouflaged injection across three model families (Claude Haiku, Llama 3.1 8B, Gemini 2.0 Flash) and three deployment domains (financial, legal, general) using 3,510 trials. Paraphrasing retrieved content before agent processing is the most consistently effective defense in this benchmark, reducing camouflage attack success rate by 55-84\% depending on model, and achieves lower attack success rates than our Llama Guard 4 configuration on every model tested. Defense effectiveness is strongly model-dependent: spotlighting halves attack success on Claude Haiku but provides no benefit on Llama 3.1 8B. Financial domain deployments face the highest residual risk at 26-33\% baseline attack success rate, with no prompting-based defense fully eliminating the threat on weaker models. These results provide the first systematic evaluation of prompting-based defenses specifically against camouflage-class injection attacks and establish benchmark-based recommendations for practitioners. All tasks use synthetically constructed professional documents; whether these benchmark rankings generalize to real enterprise documents remains an open question.

阅读与讨论 → 访问原文 →

10.

arXiv (CS.CL) 2026-06-17 DOI: arXiv:2606.17467

PARSE: Provenance-Aware Retrieval Sanitization for Professional Domain LLM Agents

作者:

Aaditya Pai↗

Prompt injection defenses evaluated on synthetic benchmarks do not generalize to real enterprise documents, which are longer, denser, and interleave legitimate authority language with factual content. We demonstrate this gap with a real-document benchmark of 122 tasks across five professional domains (financial, legal, medical, scientific, DevOps) using actual SEC filings, Federal Register rules, PubMed abstracts, arXiv papers, and GitHub postmortems. Paraphrasing, the strongest defense on synthetic benchmarks, shows no statistically significant attack success rate reduction on real documents (p=0.500) while degrading utility from 91.8% to 82.8%. We introduce PARSE (Provenance-Aware Retrieval Sanitization), a domain-aware, fact-preserving sanitization pipeline that classifies each sentence by injection likelihood, extracts structured facts before rewriting, and verifies fact preservation via a consistency-checking loop. A directiveness gate routes 59% of real enterprise documents to a lightweight path, concentrating computational cost on high-risk documents. PARSE achieves 15.6% attack success rate – a 38% reduction versus the 25.4% baseline – at 86.9% utility, the only condition that is both statistically significant (p=0.014, adequately powered) and maintains near-baseline utility. Practitioners should evaluate defenses on domain-matched real documents, not synthetic proxies.

阅读与讨论 → 访问原文 →

11.

arXiv (CS.LG) 2026-06-15 DOI: arXiv:2606.14373

Machine-learned particle flow as a foundation model for collider physics

作者:

Farouk Mokhtar↗Joosep Pata↗Michael Kagan↗Javier Duarte↗

arXiv:2606.14373v1 Announce Type: cross Abstract: The workflow from particle collision to physics analysis passes through a series of reconstruction steps that are traditionally modular and disconnected, with no shared representation linking low-level detector data to high-level analysis tasks. We show that casting event reconstruction as a machine learning problem naturally produces such a shared representation. We repurpose a machine learning model trained for particle-flow reconstruction (MLPF) to perform three distinct analysis tasks: jet flavor identification, jet energy regression, and missing momentum regression. By appending the per-particle latent representations learned during reconstruction as additional input features, we substantially improve over baselines that use kinematic features alone. We further demonstrate that a single linear layer trained using only the latent representations achieves competitive performance against state-of-the-art baseline architectures, and outperforms the baseline for missing momentum regression with approximately 35 times fewer parameters. These results demonstrate that the latent representations learned during reconstruction encode essential physics information needed for downstream analysis, establishing MLPF as a foundation model and offering a concrete step toward an end-to-end pipeline from detector data to physics analysis.

阅读与讨论 → 访问原文 →

12.

arXiv (CS.CV) 2026-06-11 DOI: arXiv:2606.11702

MedCTA: A Benchmark for Clinical Tool Agents

作者:

Tajamul Ashraf↗Hyewon Jeong↗Fida Mohammad Thoker↗Bernard Ghanem↗

To make clinically grounded decisions, medical AI agents are expected to go beyond simple recognition and be capable of tool retrieval, evidence acquisition, and integration. Existing benchmarks largely evaluate isolated perception or single-turn question answering, and therefore provide limited visibility into failures of planning, tool recruitment, and rollout reliability. We introduce MedCTA, a benchmark for evaluating medical tool agents on clinician-validated, step-implicit tasks grounded in realistic multimodal clinical inputs, including radiology images, pathology slides, and reports. MedCTA comprises 107 real-world clinical tasks with clinician-verified executable trajectories over 5 deployed tools, and supports process-aware evaluation of tool selection, argument validity, execution stability, trajectory fidelity, and outcome quality. We benchmark 18 open- and closed-source multimodal models and find that even frontier systems remain brittle in multi-step clinical tool use: autonomous rollouts are dominated by protocol failures, premature stopping, and incorrect tool recruitment, while gold-standard tool routing yields large but still incomplete gains. These results show that strong backbone perception does not translate into reliable agentic behavior in clinical settings. MedCTA provides a rigorous testbed for auditing, diagnosing, and advancing trustworthy medical AI agents. The dataset and evaluation suite are available at https://ivul-kaust.github.io/MedCTA/

阅读与讨论 → 访问原文 →

13.

arXiv (math.PR) 2026-06-17 DOI: arXiv:2606.18162

Periodicity, type $II_1$ factors and free Poisson laws in interacting Fock spaces

作者:

Vitonofrio Crismale↗Yun Gang Lu↗\'Eric Ricard↗

arXiv:2606.18162v1 Announce Type: cross Abstract: We show that the von Neumann algebra generated by position operators in a 2-periodic interacting Fock space is a type $II_1$ factor. On the probabilistic side, we prove that the squared position operators have a Marchenko-Pastur distribution with respect to the vacuum state, yielding a natural realization of free Poisson laws within this framework.

阅读与讨论 → 访问原文 →

14.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2602.08597

An Attention Mechanism for Robust Multimodal Integration in a Global Workspace Architecture

作者:

Roland Bertin-Johannet↗Lara Scipio↗Leopold Mayti\'e↗Rufin VanRullen↗

arXiv:2602.08597v3 Announce Type: replace Abstract: Robust multimodal systems must remain effective when some modalities are noisy, degraded, or unreliable. Existing multimodal fusion methods often learn modality selection jointly with representation learning, making it difficult to determine whether robustness comes from the selector itself or from full end-to-end co-adaptation. Motivated by Global Workspace Theory (GWT), we study this question using a lightweight top-down modality selector operating on top of a frozen multimodal global workspace. We evaluate our method on two multimodal datasets of increasing complexity: Simple Shapes and MM-IMDb 1.0, under structured modality corruptions. The selector improves robustness while using far fewer trainable parameters than end-to-end attention baselines, and the learned selection strategy transfers better across downstream tasks, corruption regimes, and even to a previously unseen modality. Beyond explicit corruption settings, on the MM-IMDb 1.0 benchmark, we show that the same mechanism improves the global workspace over its no-attention counterpart and yields decent benchmark performance.

阅读与讨论 → 访问原文 →

15.

arXiv (CS.CV) 2026-06-17 DOI: arXiv:2601.00215

Disentangling Perception and Reasoning in Multimodal LLMs via Reward Design

作者:

Omar Sharif↗Eftekhar Hossain↗Nikhil Singh↗Patrick Ng↗

Reinforcement learning with verifiable rewards has driven major gains in LLM reasoning, and it is intuitive to assume this recipe will transfer well to multimodal models. However, multimodal models do two things: first, perceive what is in an image, then reason about what it implies. Because these stages are graded jointly, it is hard to tell how much room reasoning alone has to grow. We study this on algorithmic visual puzzles, where both components are necessary and show that perception, not reasoning, is the binding constraint. Replacing images with simple textual descriptions raises performance by over 20 points on average for Claude models. We then evaluate six reward designs aimed at inducing visual grounding during reasoning without chain-of-thought supervision. Training Qwen-2.5-VL-7B with GRPO, reward design induces long, structured reasoning with self-reflection and visual references, yielding a 5.56-point gain over the base model. These gains are, however, uneven; no single reward improves all categories, and rewards with verifiable accuracy signals trade out-of-domain transfer for in-domain accuracy. These results point to perception-aware reward design as a path forward, so that signals correct perception at its source rather than the reasoning that inherits its errors.

阅读与讨论 → 访问原文 →

16.

arXiv (CS.CV) 2026-06-16 DOI: arXiv:2606.15779

Faithful Action-unit Causal Reasoning for Counterfactually Faithful Emotion Explanations

作者:

Van Thong Huynh↗Hong Hai Nguyen↗Thuy Pham↗Trong Nghia Nguyen↗Soo-Hyung Kim↗

Multimodal models can name the action units (AUs) behind a facial emotion, but their AU->emotion rationales are typically plausible rather than faithful: nothing forces the AUs a model invokes to be the AUs that actually drive its prediction. We cast AU->emotion reasoning as a counterfactual-consistency problem between the rationale, the label, and a structural AU->emotion causal graph G, and propose FACR, which grounds the reasoner in an independently induced, polarity-aware G and trains a counterfactual-faithfulness objective: a do-intervention on an AU that G marks causal for a class must move the prediction, while one it marks irrelevant must leave it unchanged. Faithfulness is thereby both trainable and measurable through a matching interventional metric, which we evaluate against a known causal structure, the PSPI pain-AU composition, as no existing affective-reasoning benchmark allows. We are explicit that this metric tests fidelity to the supplied structure rather than its rediscovery: it asks whether the trained reasoner invokes the AUs the structure marks causal, on held-out subjects and a second dataset. Under subject-independent evaluation on UNBC-PAIN, the objective raises the agreement between the invoked AUs and the PSPI composition from a no-objective baseline of 0.08 to 0.57, at a small detection cost; an unfaithfulness control attributes the gain to the objective. On a cross-dataset emotion transfer, the objective likewise raises fidelity to G on a seven-class task (0.50 to 0.84). Finally, we attach a language verbalizer and extend the audit to the generated text: biasing each action unit's emission by its latent activation makes the rationale faithful by construction, so that ablating an AU removes it from the explanation, a property that transfers to a second language-model backbone, whereas a freely generated rationale is unfaithful.

阅读与讨论 → 访问原文 →

17.

arXiv (CS.CL) 2026-06-11 DOI: arXiv:2606.05922

Evolving Agents in the Dark: Retrospective Harness Optimization via Self-Preference

作者:

Wenbo Pan↗Shujie Liu↗Chin-Yew Lin↗Jingying Zeng↗Xianfeng Tang↗Xiangyang Zhou↗Yan Lu↗Xiaohua Jia↗

AI agents rely on a harness of skills, tools, and workflows to solve complex problems. Continually improving this harness is essential for adapting to new tasks. However, existing optimization methods typically require ground-truth validation sets, yet such labeled data is difficult to acquire in practical deployment settings. To address this problem, we introduce Retrospective Harness Optimization (RHO), a self-supervised method that optimizes the agent harness using only past trajectories. Specifically, RHO selects a diverse coreset of challenging tasks from past trajectories and re-solves them in parallel. The agent analyzes these rollouts using self-validation and self-consistency, then generates candidate harness updates and selects the most effective one by its own pairwise self-preference. We evaluate RHO across three diverse domains, spanning software engineering, technical work, and knowledge work. Notably, a single optimization round improves the pass rate on SWE-Bench Pro from 59% to 78% without any external grading. Furthermore, our analysis demonstrates that RHO effectively targets prior failure modes. As a result, the optimized harness alters the agent's behavior patterns and sustains higher accuracy during long-horizon sessions.

阅读与讨论 → 访问原文 →

18.

arXiv (CS.AI) 2026-06-15 DOI: arXiv:2606.14031

Applicability Condition Extraction for Therapeutic Drug-Disease Relations

作者:

Guanting Luo↗Noriki Nishida↗Yuji Matsumoto↗Yuki Arase↗

arXiv:2606.14031v1 Announce Type: new Abstract: Identifying conditions that a certain drug takes therapeutic effect on a target disease is crucial for clinical decision-making support. However, most existing biomedical information extraction methods have focused on identifying only relations between drugs and diseases, while largely overlooking the context-specific conditions where such relations can apply. To address this problem, we introduce the task of applicability condition extraction for therapeutic drug–disease relations from biomedical research literature. We create the first dataset that has manually annotated triples of drugs, diseases, and applicability conditions on biomedical paper abstracts with 1,119 drug-disease pairs. Using this dataset, we systematically evaluate the performance of a range of existing methods. In addition, we propose a new method that enhances LoRA to consider relations between drugs and diseases. Our method consistently outperforms strong baselines across different evaluation settings. The source code and dataset of this paper can be obtained from: https://github.com/guantingluo98/Drug-ACE

阅读与讨论 → 访问原文 →

19.

arXiv (CS.AI) 2026-06-19 DOI: arXiv:2502.19193

Simulation of Language Evolution under Regulated Social Media Platforms: A Synergistic Approach of Large Language Models and Genetic Algorithms

作者:

Jinyu Cai↗Yusei Ishimizu↗Mingyue Zhang↗Munan Li↗Jialong Li↗Kenji Tei↗

arXiv:2502.19193v2 Announce Type: replace-cross Abstract: Social media platforms frequently impose restrictive policies to moderate user content, prompting the emergence of creative evasion language strategies. This paper presents a multi-agent framework based on Large Language Models (LLMs) to simulate the iterative evolution of language strategies under regulatory constraints. In this framework, participant agents, as social media users, continuously evolve their language expression, while supervisory agents emulate platform-level regulation by assessing policy violations. To achieve a more faithful simulation, we employ a dual design of language strategies (constraint and expression) to differentiate conflicting goals and utilize an LLM-driven GA (Genetic Algorithm) for the selection, mutation, and crossover of language strategies. The framework is evaluated using two distinct scenarios: an abstract password game and a realistic simulated illegal pet trade scenario. Experimental results demonstrate that as the number of dialogue rounds increases, both the number of uninterrupted dialogue turns and the accuracy of information transmission improve significantly. Furthermore, a user study with 40 participants validates the real-world relevance of the generated dialogues and strategies. Moreover, ablation studies validate the importance of the GA, emphasizing its contribution to long-term adaptability and improved overall results.

阅读与讨论 → 访问原文 →

20.

arXiv (CS.CV) 2026-06-16 DOI: arXiv:2605.10157

MolSight: Molecular Property Prediction with Images

作者:

Aaditya Baranwal↗Akshaj Gupta↗Yogesh S Rawat↗Shruti Vyas↗

Every molecule ever synthesised can be drawn as a 2D skeletal diagram, yet in modern property prediction this universally available representation has received less focus in favour of molecular graphs, 3D conformers, or billion-parameter language models, each imposing its own computational and data-engineering overhead. We present $MolSight$, the first systematic large-scale study of vision-based Molecular Property Prediction (MPP). Using 10 vision architectures, 7 pre-training strategies, and $2\,M$ molecule images, we evaluate performance across 10 downstream tasks spanning physical-property regression, drug-discovery classification, and quantum-chemistry prediction. To account for the wide variation in structural complexity across pre-training molecules, we further propose a $chemistry-informed curriculum$: five structural complexity descriptors partition the corpus into five tiers of increasing chemical difficulty, consistently outperforming non-curriculum baselines. We show that a single rendered bond-line image, processed by a vision encoder, is sufficient for competitive molecular property prediction, i.e. $chemical insight from sight alone$. The best curriculum-trained configuration achieves the top result on $5 of 10$ benchmarks and top two on $all 10$, at $$80$\times$ lower$$ FLOPs than the nearest multi-modal competitor.

阅读与讨论 → 访问原文 →

21.

arXiv (CS.CL) 2026-06-18 DOI: arXiv:2606.19170

Dango: A Strictly L1-Only Large Language Model for Studying Second Language Acquisition

作者:

Shiho Matta↗Yin Jou Huang↗Fei Cheng↗Takashi Kodama↗Hirokazu Kiyomaru↗Yugo Murawaki↗

We introduce Dango, a 1.8B-parameter large language model designed for controlled studies of L1-to-L2 (Japanese-to-English) transfer in second language acquisition (SLA). While previous studies have explored SLA in language models, they have predominantly relied on smaller or non-decoder models, limiting their ability to generate open-ended text and reducing their suitability as practical L2 simulators. We identify a key challenge when scaling models to this size: L2 contamination within the "monolingual" pretraining corpus used for L1 acquisition. To address this, we propose a filtering method to reduce premature exposure to English while preserving realistic, minimal exposure. We then fine-tune the model on LLM-generated L2-learning lessons to simulate the L2 acquisition process. Our evaluations confirm that Dango develops human-like L2 production patterns, outperforming both unfiltered and standard multilingual baselines. We release the model, data, and code to facilitate reproducible computational SLA research and learner-facing applications.

阅读与讨论 → 访问原文 →

22.

arXiv (CS.CL) 2026-06-12 DOI: arXiv:2606.05405

Agents' Last Exam

作者:

Yiyou Sun↗Xinyang Han↗Weichen Zhang↗Yuanbo Pang↗Tianyu Wang↗Yuhan Cao↗Yixiao Huang↗Chris Duroiu↗Haoyun Zhang↗Jeffrey Lin↗Weishu Zhang↗Tyler Zeng↗…

Recent AI systems have achieved strong results on a wide range of benchmarks, yet these gains have not translated into economically meaningful deployment across many professional domains. We argue that this gap is largely an evaluation problem: widely used benchmarks lack sustained performance measurement on real and economically valuable workflows. This paper introduces Agents' Last Exam (ALE), a benchmark designed to evaluate AI agents on long horizon, economically valuable, real world tasks with verifiable outcomes. Developed in collaboration with 250+ industry experts, ALE covers non-physical industries defined with reference to O*NET / SOC 2018 (the U.S. federal occupational taxonomy). It is organized around a task taxonomy with 55 sub fields grouped into 13 industry clusters covering 1K+ tasks. Current results show that the hardest tier remains far from saturated: across mainstream harness and backbone configurations, the average full pass rate is below 1%. ALE is designed as a living benchmark: its task pool grows continuously as new workflows and industries are onboarded. More broadly, ALE is intended not merely as another leaderboard, but as an instrument for closing the gap between benchmark success and GDP relevant impact.

阅读与讨论 → 访问原文 →

23.

arXiv (CS.CV) 2026-06-12 DOI: arXiv:2606.13345

JointEdit3D: Feed-Forward 3D Scene Editing in a Unified Latent Space

作者:

Xinnan Zhu↗Ruijie Xu↗Jiayu Ying↗Daoguo Dong↗Jiachen Xu↗Yuan Xie↗Xin Tan↗

Existing 3D scene editing methods typically rely on per-scene optimization over explicit 3D representations or cascaded edit-and-reconstruct pipelines, resulting in high test-time cost, limited 3D awareness, and structural inconsistencies. To couple appearance synthesis and geometry prediction during editing, we build on a unified RGB-geometry reconstruction-generation latent space and adapt it to feed-forward 3D scene editing. The resulting framework, JointEdit3D, performs asymmetric latent inpainting by observing only a single edited RGB reference latent and generating the remaining RGB views and edited geometry latent under source-scene anchoring. JointEdit3D introduces a dedicated SceneAnchor Branch to inject source-scene structure without forcing direct copying, and adopts edit/background-aware losses to balance edited-region fidelity with unedited-content preservation. To address the lack of paired resources for standardized 3D scene editing evaluation, we introduce SceneEdit3D-15K, a dataset with 15K paired editing samples and renderer-provided 3D annotations, together with SceneEdit3D-Bench, a curated 100-sample benchmark. Experiments show that JointEdit3D improves edited-region quality and 3D structural completeness over prior baselines while maintaining competitive background preservation.

阅读与讨论 → 访问原文 →

24.

medRxiv (Medicine) 2026-06-22 DOI: HASH:d7d811ef71b29bb4186c3590d8acc97a

Cumulative Metabolic Exposure to Hyperglycemia and Risk of Cardiovascular and Limb Events in Peripheral Artery Disease

作者:

James↗Li↗Somisetty↗Nguyen↗Vaughan-Sarrazin↗M. S↗Lund↗Tsai↗Smolderen↗K. G↗Hoffman↗R. M↗…

Background: Although diabetes is a potent risk factor for the development of peripheral artery disease (PAD), the effect of cumulative metabolic exposure to hyperglycemia on risk of cardiovascular or limb events in patients with PAD remains unclear. Methods: The Peripheral Artery Disease: Long-term Survival (PEARLS) is a longitudinal registry of Veterans with newly diagnosed PAD identified using a natural language processing approach. Included patients had ankle brachial index [≤]0.9 or toe brachial index [≤]0.7, and no history of lower extremity revascularization or major amputation. Among patients with diabetes in this cohort, we assessed cumulative exposure to hyperglycema based on a 24-month rolling average of hemoglobin (Hgb) A1c values, categorized as [≤]7%, >7% to [≤]8%, and >8%. Multivariable Cox regression models evaluated the association between categories of HgbA1c, modeled as a time-varying exposure, and risk of cardiovascular (CV: myocardial infarction or stroke) and limb (chronic limb threatening ischemia [CLTI] or major amputation) events. Results: Among 45,109 patients with new diagnosis of PAD and pre-existing diabetes, the mean HgbA1c at baseline was 7.5%, with nearly one-third (30.4%) having HgbA1c >8%. The mean age was 70.4 years, 19.8% were Black and 4% were Hispanic. Patients with baseline HgbA1c >8% were younger and compared to those with HgbA1c [≤]7%, more likely to have coronary disease, kidney disease, and obesity. Over a median follow up of 4.2 years, 8,306 (18.4%) patients experienced a CV event, and 8,199 (18.2%) experienced a limb event. The adjusted association between HgbA1c and hazard of CV events was 12% higher in patients exposed to HgbA1c >7% to [≤]8% (HR 1.12; 95%CI: 1.05-1.18) and 38% higher in those exposed to HgbA1c >8% (HR 1.38; 95%CI: 1.30-1.46), compared to HgbA1c 7% to [≤]8% (HR 1.20; 95%CI: 1.13-1.28) and HgbA1c >8% (HR 1.60; 95%CI: 1.51-1.70), respectively when compared to HgbA1c [≤]7%. These findings were consistent in subgroups based on age and severity of PAD. Conclusions: Among diabetic patients with PAD, cumulatiave metabolic exposure to hyperglycemia is associated with a markedly increased risk of clinical events, especially limb events.

阅读与讨论 → 访问原文 →

25.

arXiv (CS.AI) 2026-06-12 DOI: arXiv:2601.19072

HalluJudge: A Reference-Free Hallucination Detection for Context Misalignment in Code Review Automation

作者:

Kla Tantithamthavorn↗Hong Yi Lin↗Patanamon Thongtanunam↗Wachiraphan Charoenwet↗Minwoo Jeong↗Ming Wu↗

arXiv:2601.19072v3 Announce Type: replace-cross Abstract: Large Language models (LLMs) have shown strong capabilities in code review automation, such as review comment generation, yet they suffer from hallucinations – where the generated review comments are ungrounded in the actual code – poses a significant challenge to the adoption of LLMs in code review workflows. To address this, we explore effective and scalable methods for a hallucination detection in LLM-generated code review comments without the reference. In this work, we design HalluJudge that aims to assess the grounding of generated review comments based on the context alignment. HalluJudge includes four key strategies ranging from direct assessment to structured multi-branch reasoning (e.g., Tree-of-Thoughts). We conduct a comprehensive evaluation of these assessment strategies across Atlassian's enterprise-scale software projects to examine the effectiveness and cost-efficiency of HalluJudge. Furthermore, we analyze the alignment between HalluJudge's judgment and developer preference of the actual LLM-generated code review comments in the real-world production. Our results show that the hallucination assessment in HalluJudge is cost-effective with an F1 score of 0.85 and an average cost of $0.009. On average, 67% of the HalluJudge assessments are aligned with the developer preference of the actual LLM-generated review comments in the online production. Our results suggest that HalluJudge can serve as a practical safeguard to reduce developers' exposure to hallucinated comments, fostering trust in AI-assisted code reviews.

阅读与讨论 → 访问原文 →