论文检索

Kewei Chen, Yayu Long, Shuai Li, Mingsheng Shang

Addressing the escalating security vulnerabilities in Vision-Language-Action (VLA) models, this study investigates backdoor attacks targeting the visual pathway. We identify a core obstacle causing the failure of traditional attack paradigms: "Gradient Interference." This phenomenon represents an optimization failure triggered by conflicting strategies during end-to-end training. To resolve this, we propose an Adaptive Threat-Aware Adversarial Tuning (ATAAT) framework. Through its core "Threat-Method Adaptive Mapping" mechanism, ATAAT intelligently selects the optimal gradient decoupling strategy based on the adversary’s capabilities. Extensive experiments demonstrate that ATAAT exhibits significant advantages, achieving a highly robust Targeted Attack Success Rate (TASR > 80%) while maintaining extreme stealthiness with merely a 5% poisoning rate. It efficiently handles complex semantic-level triggers and achieves implicit decoupled attacks in data poisoning scenarios for the first time. This work reveals a critical security vulnerability in VLAs and provides theoretical and methodological support for future defense architectures.

Pengxiang Liu, Tao Ren, Wei Xiong, Tingrui Yang, Junjie Wang, Jun HU

Large Language Models (LLMs) have shown impressive reasoning capabilities in agents for complex interactive environments. However, these agents often suffer from hallucinations and lack grounding, leading to unreliable actions that conflict with real-world constraints. Existing approaches mitigate some issues through implicit imitation or sparse reinforcement learning but rely on fitting data distributions without explicitly understanding environmental constraints, often generating actions that are behaviorally distorted or environmentally impermissible. To address this, we introduce OntoGuard, an ontological framework designed to guard LLM agents by enforcing environmental and behavioral admissibility. These constraints are constructed by extracting knowledge from oracle demonstrations, supplemented with world knowledge inherent in LLMs and general knowledge bases. During inference, OntoGuard functions as an active interceptor, using a graph-based constraint-checking mechanism to reject invalid actions and prompt self-correction before acting. Experiments on both ScienceWorld and VirtualHome demonstrate OntoGuard’s advantage over state-of-the-art methods, validating its ability to enforce physical and behavioral constraints while preventing invalid actions.

Xiaoyun Zhang, Xiaojian Yuan, Di Huang, Wang You, Chen Hu, Jingqing Ruan, Kejiang Chen, Xing Hu

Reasoning ability has become a defining capability of Large Language Models (LLMs), with Reinforcement Learning with Verifiable Rewards (RLVR) emerging as a key paradigm to enhance it. However, RLVR training often suffers from policy entropy collapse, where the policy becomes overly deterministic, hindering exploration and limiting reasoning performance. While entropy regularization is a common remedy, its effectiveness is highly sensitive to the fixed coefficient, making it unstable across tasks and models. In this work, we revisit entropy regularization in RLVR and argue that its potential has been largely underestimated. Our analysis shows that (i) tasks of varying difficulty demand distinct exploration intensities, and (ii) balanced exploration may require the policy entropy to be maintained within a moderate range below its initial level. Therefore, we propose Adaptive Entropy Regularization (AER) — a framework that dynamically balances exploration and exploitation via three components: difficulty-aware coefficient allocation, initial-anchored target entropy, and dynamic global coefficient adjustment. Experiments on multiple mathematical reasoning benchmarks show that AER consistently outperforms baselines, improving both reasoning accuracy and exploration capability. Codes are available at https://anonymous.4open.science/r/AER-ACL .

Fulong Fan, Peilin Liu, Liu FengZhe, Shuyan Yang, Gang Yan

Large language models perform well on many reasoning tasks, yet they often lack awareness of whether their current knowledge or reasoning state is complete. In non-interactive puzzle settings, the narrative is fixed and the underlying structure is hidden; once a model forms an early hypothesis under incomplete premises, it can propagate that error throughout the reasoning process, leading to unstable conclusions. To address this issue, we propose SABA, a reasoning framework that explicitly introduces self-awareness of missing premises before making the final decision. SABA formulates reasoning as a recursive process that alternates between structured state construction and obstacle resolution: it first applies Information Fusion to consolidate the narrative into a verifiable base state, and then uses Query-driven Structured Reasoning to identify and resolve missing or underspecified premises by turning them into queries and progressively completing the reasoning state through hypothesis construction and state refinement. Across multiple evaluation metrics, SABA achieves the best performance on all three difficulty splits of the non-interactive Detective Puzzle benchmark, and it also maintains leading results on multiple public benchmarks.

Zihan Chen, Yiming Zhang, Hengguang Zhou, Zenghui Ding, Yining Sun, Cho-Jui Hsieh

Current benchmarks are inadequate for evaluating progress in reinforcement learning (RL) for large language models (LLMs). Despite recent benchmark gains reported for RL, we find that training on these benchmarks’ training sets achieves nearly the same performance as training directly on the test sets, suggesting that the benchmarks cannot reliably separate further progress. To study this phenomenon, we introduce a diagnostic suite and the Oracle Performance Gap (OPG) metric that quantifies the performance difference between training on the train split versus the test split of a benchmark. We further analyze this phenomenon with stress tests and find that, despite strong benchmark scores, existing RL methods struggle to generalize across distribution shifts, varying levels of difficulty, and counterfactual scenarios: shortcomings that current benchmarks fail to reveal. We conclude that current benchmarks are insufficient for evaluating generalization and propose three core principles for designing more faithful benchmarks: sufficient difficulty, balanced evaluation, and distributional robustness.

Zaiyuan Di, Jianting Chen, Yunxiao Yang, Xiaoying Gao, Li Yang, Zhihao Wang, Yang Xiang

While recent studies have increasingly emphasized the role of reflection in code repair tasks, existing benchmarks still target the repair generation capability of LLMs, lacking fine-grained evaluation of reflection generation capability. To this end, we propose Code Reffix, a benchmark featuring an automated pipeline with oracle reflections and a dual-task protocol to decouple the evaluation of reflection from repair. Through extensive experiments on 14 LLMs and fine-tuning analysis, we aim to pinpoint performance bottlenecks of code repair, quantify reflection quality, and verify the value of reflection optimization. Evaluations reveal that underperforming reflection capabilities of small-scale LLMs remain a major bottleneck for code repair. By quantifying this gap, Code Reffix provides a critical foundation for optimizing LLMs to achieve superior repair performance.

Bajian Xiang, Tingwei Guo, Xuan Chen, Yang Han

Large Speech Language Models (LSLMs) typically operate at high token rates (tokens/s) to ensure acoustic fidelity, yet this results in sequence lengths that far exceed the underlying semantic content, incurring prohibitive inference costs. In this paper, we empirically revisit the necessity of such granular token-level processing. Through layer-wise oracle interventions, we unveil a structured redundancy hierarchy: while shallow layers encode essential acoustic details, deep layers exhibit extreme redundancy, allowing for aggressive compression. Motivated by these findings, we introduce Affinity Pooling, a training-free, similarity-based token merging mechanism. By strategically applying this method at both input and deep layers, we effectively compress speech representations without compromising semantic information. Extensive evaluations across three tasks demonstrate that our approach reduces prefilling FLOPs by 27.48% while maintaining competitive accuracy. Practical deployment further confirms significant efficiency gains, yielding up to \sim1.7\times memory savings and \sim1.1\times faster time-to-first-token on long utterances. Our results challenge the necessity of fully distinct token representations, providing new perspectives on LSLM efficiency.

Hongbin Na, Zimu Wang, Zhaoming Chen, Peilin Zhou, Yining Hua, Grace Ziqi Zhou, Haiyang Zhang, Tao Shen, Wei Wang, John Torous 等

Psychological defenses are strategies, often automatic, that people use to manage distress. Rigid use or overuse of defenses is negatively linked to mental health and shapes what speakers disclose and how they accept or resist help. However, defenses are complex and difficult to reliably measure, particularly in clinical dialogues. We introduce PsyDefConv, a dialogue corpus with help seeker utterances labeled for defense level, and DMRS Co-Pilot, a four-stage pipeline that provides evidence-based pre-annotations. The corpus contains 200 dialogues and 4,709 utterances, including 2,336 help seeker turns, with double-blind labeling reaching Cohen’s kappa of 0.639. In a counterbalanced study, the co-pilot reduced average annotation time by 24.0%. In expert review, it averaged 4.62 for evidence supportiveness, 4.44 for clinical plausibility, and 4.40 for insight on a seven-point scale. Benchmarks with strong large language models (LLMs) in zero-shot and fine-tuning settings demonstrate clear headroom, with the best macro F1-score around 30% and a tendency to overpredict mature defenses. Corpus analyses confirm that mature defenses are most common and reveal emotion-specific deviations. We release the corpus, annotations, code, and prompts to support research on defensive functioning in language.

Haonan Bian, Yutao Qi, Rui Yang, Yuanxi Che, Jiaqian Wang, Heming Xia, Ranran Zhen

Large Language Models (LLMs), despite their success in question answering, exhibit limitations in complex multi-hop question answering (MQA) tasks that necessitate non-linear, structured reasoning. This limitation stems from their inability to adequately capture deep conceptual relationships between entities. To overcome this challenge, we present ORACLE (Ontology-driven Reasoning And Chain for Logical Elucidation), a training-free framework that combines LLMs’ generative capabilities with the structural benefits of knowledge graphs. Our approach operates through three stages: (1) dynamic construction of question-specific knowledge ontologies using LLMs, (2) transformation of these ontologies into First-Order Logic (FOL) reasoning chains, and (3) systematic decomposition of the original query into logically coherent sub-questions. Extensive experiments across a diverse set of models and standard MQA benchmarks demonstrate that our framework achieves competitive performance while producing more interpretable reasoning chains.

Jieran Li, Xiuyuan Hu, Yang Zhao, Dongbiao Sun, Hao Zhang

Despite their strong generative capabilities, large language models frequently exhibit hallucinations, particularly due to outside-boundary confidence where incorrect assertions are produced with high statistical certainty. Existing approaches commonly use output probability as a proxy for truthfulness; however, this signal is confounded by epistemic uncertainty and cannot reliably distinguish genuine uncertainty from fabricated content. We argue that effective hallucination detection requires integrating surface-level confidence with signals that reflect the model’s underlying epistemic state. To this end, we propose Answer-level Intrinsic Cognition (AIC), a model-agnostic metric that captures epistemic boundary deviations by measuring answer-level stability across multiple stochastic forward passes. By coupling AIC with conventional output uncertainty, we derive a composite metric that disentangles within-boundary uncertainty from outside-boundary confidence. Across three public question-answering benchmarks and diverse model scales, the two-dimensional score consistently outperforms strong uncertainty-only baselines, with larger gains on adversarially constructed hallucination sets. The code is available at: https://github.com/HXYfighter/AIC-ACL2026.

Juntuo Wang, Yuming Qiao, Yifan Yang, Lunxi Yuan, Liang Luo, Dan Meng

Vision-language models (VLMs) are increasingly adopted as judges for subjective assessment, yet absolute scoring remains brittle due to inconsistent scales and inherent preference biases. To bridge this gap, we propose \text{S}^2\text{AD} (**Semantic-Anchored Scale-Agnostic Distillation**), a novel easy-to-hard framework that operationalizes subjective assessment as comparative analysis, conceptualizing the judge’s evolution from mimesis to metamorphosis. In Stage 1 (Mimesis), we introduce Dynamic Soft Positioning (DSP) to train the judge to compare a query against retrieved reference images, establishing a relative evaluation space that ensures consistent ordering under heterogeneous scales. In Stage 2 (Metamorphosis), this comparative capability is internalized via Language Buttons—discrete semantic levels serving as a retrieval-free internal reference. Optimized with Group Relative Policy Optimization (GRPO), \text{S}^2\text{AD} achieves efficient, scale-steerable inference that adapts to diverse grading standards. Our framework reaches state-of-the-art performance across multiple benchmarks, validating the effectiveness of internalized comparative priors for robust, rank-invariant, and scale-steerable evaluation. The code is available at: https://github.com/SpatialVision-Research/SSAD_ACL2026_Findings.

Jiyang Zheng, Islam Nassar, Thanh Vu, Xu Zhong, Yang Lin, Tongliang Liu, Long Duong, Yuan-Fang Li

Medical coding converts free-text clinical notes into standardized diagnostic and procedural codes, which are essential for billing, hospital operations, and medical research. Unlike ordinary text classification, it requires multi-step reasoning: extracting diagnostic concepts, applying guideline constraints, mapping to hierarchical codebooks, and ensuring cross-document consistency. Recent advances leverage agentic LLMs, but most rely on rigid, manually crafted workflows that fail to capture the nuance and variability of real-world documentation, leaving open the question of how to systematically learn effective workflows. We present MedDCR, a closed-loop framework that treats workflow design as a learning problem. A Designer proposes workflows, a Coder executes them, and a Reflector evaluates predictions and provides constructive feedback, while a memory archive preserves prior designs for reuse and iterative refinement. On benchmark datasets, MedDCR outperforms state-of-the-art baselines and produces interpretable, adaptable workflows that better reflect real coding practice, improving both the reliability and trustworthiness of automated systems.

Qianen Zhang, Zeyu Yang, Satoshi Nakamura

Simultaneous Machine Translation (SiMT) requires high-quality translations under strict real-time constraints, which traditional policies with only READ/WRITE actions cannot fully address. We extend the action space of SiMT with four adaptive actions: **Sentence_Cut**, **Drop**, **Partial_Summarization** and **Pronominalization**, which enable real-time restructuring, omission, and simplification while preserving semantic fidelity. We adapt these actions in a large language model (LLM) framework and construct training references through action-aware prompting. To evaluate both quality and word-level monotonicity, we further develop a latency-aware TTS pipeline that maps textual outputs to speech with realistic timing. Experiments on the ACL60/60 English-Chinese, English-German and English-Japanese benchmarks show that our framework consistently improves semantic metrics and achieves lower delay compared to reference translations and salami-based baselines. Notably, combining **Drop** and **Sentence_Cut** leads to consistent improvements in the balance between fluency and latency. These results demonstrate that enriching the action space of LLM-based SiMT provides a promising direction for bridging the gap between human and machine interpretation.

Ashutosh Hathidara, Julien Yu, Sebastian Schreiber

Large language models (LLMs) are increasingly tasked with invoking enterprise APIs, yet they routinely falter when near-duplicate tools vie for the same user intent or when required arguments are left underspecified. We introduce **DiaFORGE** (**Dia**logue **F**ramework for **O**rganic **R**esponse **G**eneration **E**valuation), a disambiguation-centric, three-stage pipeline that (i) synthesizes persona-driven, multi-turn dialogues in which the assistant must distinguish among highly similar tools, (ii) performs supervised fine-tuning of open-source models with reasoning traces across 3B - 70B parameters, and (iii) evaluates real-world readiness via a dynamic suite that redeploys each model in a live agentic loop and reports end-to-end goal completion alongside conventional static metrics. On our dynamic benchmark DiaBENCH, models trained with DiaFORGE raise tool-invocation success by **27 pp over GPT-4o** and by **49 pp over Claude-3.5-Sonnet**, both under optimized prompting. To spur further research, we release an open corpus of **5000 production-grade enterprise API** specifications paired with rigorously validated, disambiguation-focused dialogues, offering a practical blueprint for building reliable, enterprise-ready tool-calling agents.

Zhihan Zhou, Daqian Shi, Lida Shi, Rui Song, Peiqiang Qiu, Xiaolei Diao, Hao Xu

Research on ancient Chinese language is of great significance for tracing Chinese history and civilization. In the field of large language models, studies on the pre-Qin excavated documents such as Oracle Bone Inscriptions, Bronze Inscriptions, and Bamboo Book of Chu remain insufficient. This is because these ancient characters have a low level of digitization, training corpora are extremely scarce, and they typically contain complex and rich semantic information. Therefore, we propose an ancient character semantic-aware embedding for large language models. This embedding integrates both the glyph and lexicality of ancient characters and maps them to the modern Chinese semantic space. We also design a two-stage method for lightweight and parameter-efficient training of the embedding. Finally, we conduct extensive experiments on excavated documents from the pre-Qin period, and the results demonstrate the effectiveness of our approach.

Xuanming Zhang, Shwan Ashrafi, Aziza Mirsaidova, Amir H. Rezaeian, Miguel Ballesteros, Lydia Chilton, Zhou Yu, Dan Roth

We study the reasoning behavior of large language models (LLMs) under limited computation budgets. In such settings, producing useful partial solutions quickly is often more practical than exhaustive reasoning, which incurs high inference costs. Many real-world tasks, such as trip planning, require models to deliver the best possible output within a fixed reasoning budget. We introduce an anytime reasoning framework and the Anytime Index, a metric that quantifies how effectively solution quality improves as reasoning tokens increase. To further enhance efficiency, we propose an inference-time self-improvement method using LLM-synthesized preference data, where models learn from their own reasoning comparisons to produce better intermediate solutions. Experiments on NaturalPlan (Trip), AIME, and GPQA datasets show consistent gains across Grok-3, GPT-oss, GPT-4.1/4o, and LLaMA models, improving both reasoning quality and efficiency under budget constraints.

Woojin Lee, Jin-Xia Huang

State-of-the-art code generation frameworks rely on mental simulation, where LLMs internally trace execution to verify correctness. We expose a fundamental limitation: the Mental-Reality Gap—where models hallucinate execution traces and confidently validate buggy code. This gap manifests along two orthogonal dimensions: the Specification Gap (overlooking edge cases during planning) and the Verification Gap (hallucinating correct behavior for flawed code). We propose SolidCoder with a simple principle: don’t imagine—execute. The S.O.L.I.D. architecture addresses both dimensions by forcing edge-case awareness before algorithm design and replacing imagined traces with sandboxed execution using property-based oracles. With GPT-4o, SolidCoder achieves state-of-the-art pass@1 performance: 95.7% on HumanEval (+0.6%p), 77.0% on CodeContests (+4.3%p), and 26.7% on APPS (+3.4%p). Ablation reveals that edge-case awareness provides the largest individual gain, while execution grounding catches categorically different errors that specification improvements cannot address. These gains generalize to RL post-trained models, validating that bridging both gap dimensions is essential for robust code synthesis. We release our code and framework to facilitate future research.

Yinghao Chen, Wantong Xie, Shuli Zeng, Sijia Zhang, Xiaotian Pan, Feng Wu, Xiangyang Li

Large language models (LLMs) and vision-language models (VLMs) are increasingly used as optimization assistants to produce solutions, generate solver-executable programs, or both. However, current evaluations are misaligned with deployment in three ways: they (P1) fail to represent multimodal problem specifications, (P2) score outcomes only and cannot localize where failures occur along the modeling pipeline, and (P3) rarely report inference cost, obscuring reliability–cost trade-offs. We introduce Graph Optimization benchmark (GOBench), an aligned multimodal benchmark with solver-derived oracles and a four-layer diagnostic protocol that evaluates intermediate artifacts as well as end results, together with the Visual Inference Penalty (VIP) to measure multimodal overhead. Across frontier and open-weight models under paired text-only vs. T+V settings, we find that vision reliably increases inference cost, while its reliability impact is regime-dependent: frontier models often benefit from visual grounding, whereas several mid-tier/open models exhibit a Visual Paradox where vision reduces downstream executability and verification coverage. End-to-end success is frequently bottlenecked by intermediate-stage dropout; supervised fine-tuning on intermediate targets can mitigate this attrition in open models, enabling a reproducible harness for diagnosing failure modes and quantifying reliability–cost trade-offs.

Huopu Zhang, Yanguang Liu, Miao Zhang, Zirui He, Mengnan Du

Predicting earnings surprises through the analysis of earnings conference call transcripts has attracted increasing attention from the financial research community. Conference calls serve as critical communication channels between company executives, analysts, and shareholders, offering valuable forward-looking information. However, these transcripts present significant analytical challenges, typically containing over 5,000 words with substantial redundancy and industry-specific terminology that creates obstacles for language models. In this work, we propose the Sparse Autoencoder for Financial Representation Enhancement (SAE-FiRE) framework to address these limitations by extracting key information while eliminating redundancy. SAE-FiRE employs Sparse Autoencoders (SAEs) to efficiently identify patterns and filter out noises, and focusing specifically on capturing nuanced financial signals that have predictive power for earnings surprises. Experimental results indicate that the proposed method can significantly outperform comparing baselines.

A Pranav, Shane Storks, Anne Lauscher

ACL removed the anonymity period for conference submissions in February 2024, allowing unrestricted preprinting during review.To examine how preprints and author recognition affect outcomes across institutional hierarchies, we track preprinting trends for 47k publications, survey 75 NLP researchers, interview 14 community members, and analyze 1.9k peer reviews. We observe that more elite institutions post preprints more frequently (52% vs. 36% by 2025). Most participants agree that preprinting gives these institutions an advantage in peer review, and indeed, reviewer knowledge of authors inflates scores at elite institutions (d = 0.43, p < 0.001) but not elsewhere, also lowering review quality. Nonetheless, the anonymity period was found largely ineffective; instead, underrepresented researchers emphasize struggles with visibility, review quality, and external structural barriers. To counteract these inequities, we make recommendations for review quality improvement and increasing investment in diversity initiatives that center the perspectives of affected communities.