Large language model-based agents increasingly rely on external skills to reuse task-specific knowledge without retraining the model. As skills become common for extending agent capabilities, understanding creation and evolution is critical for reliable agent systems. However, existing studies mainly examine agent skills from a static perspective, leaving their evolution in real-world repositories largely unexplored. This limits our understanding of the patterns and motivations behind skill changes. To fill this gap, we present a large-scale empirical study of agent skills, capturing 84,200 skills and 750,898 commits from GitHub. We manually construct taxonomies of skill components, change patterns, and evolution motivations, and build an automated annotation pipeline to label changes at scale. Based on the taxonomy, we conduct quantitative analyses of skill composition and post-creation evolution frequency, together with qualitative assessments of modified components, change patterns, and motivations. Our findings reveal five implications: (i) reliability receives less attention in skill composition and change, motivating execution-contract skill design that incorporates verification and recovery mechanisms; (ii) evolution changes concentrate in project-coupled skills, calling for automated just-in-time skill updating mechanisms; (iii) change patterns are dominated by incremental enrichment and refinement, motivating pattern-guided skill enhancers that diagnose and improve incomplete skill components; (iv) redundancy is a practical maintenance issue, highlighting the need for context-efficient skill representation; (v) security-motivated changes show limited risk reduction, suggesting skill regression testing methods that objectively verify skill functionality and security. Beyond these insights, our dataset provides a basis for future studies on automated skill techniques.
Hao Jiang, Peiru Du, Pengfei Yao, Mengting Li, Siyuan Lou, Kuo Cai, Sheng Yu, Qiang Luo 等 13 位作者
Large language models (LLMs) have demonstrated strong reasoning capabilities, motivating their use as the backbone of foundation recommendation models (FRMs). Existing methods enhance recommendations through explicit Chain-of-Thought (CoT) reasoning under a Think-then-Answer paradigm. However, explicit CoT incurs substantial inference overhead by generating lengthy reasoning traces and relies on manually designed templates that struggle to capture diverse, dynamic user interests. We propose OneLatent, an efficient latent reasoning framework that compresses explicit reasoning traces into several learnable latent tokens, enabling Latent-Reason-then-Answer inference without generating verbose traces. OneLatent first introduces Multi-View Adaptive CoT (MV-ACoT), which creates diverse, high-quality teacher-generated supervision by exploring user interests from multiple perspectives and automatically adapting reasoning complexity to each instance. Building on pretrained FRMs, it then uses a three-stage latent-token alignment paradigm to progressively internalize CoT traces into learnable latent tokens. Finally, a multistage curriculum-based post-training strategy activates latent-token reasoning for downstream recommendation tasks. Experiments on an industrial-scale Kuaishou dataset and the public Kuaishou LLM-Rec benchmark show that OneLatent consistently outperforms explicit CoT-based methods and traditional baselines. Compared with the Think and No-Think variants of FRMs, OneLatent improves SID@64 by 17.44% and 9.33%, respectively, while achieving over 17x higher online inference throughput. We further develop a production serving system for scalable, real-time FRM inference. An online A/B test in Kuaishou's local-services advertising scenario shows that deploying OneLatent with this system yields an estimated 9.6% revenue lift over strong online baselines, including OneRec and OneReason.
Fluid antenna systems (FASs) have shown that activating a single port from a densely packed antenna array can yield spatial diversity comparable to that of tens of independent antennas, a property that has substantially improved communication reliability and efficiency. The fluid antenna array (FAA) extends this concept to multi-port configurations while avoiding complex multi-dimensional spatial correlation, and offers significant geometrical flexibility under a finite aperture and a limited port budget. These attributes motivate environment-adaptive array geometry for sensing in integrated sensing and communication (ISAC) systems. Although this flexibility enables adaptive array geometries that exploit near-field distance-angle information, improving multi-parameter target estimation under communication and deployment constraints remains challenging. In this paper, we propose a planar FAA-assisted near-field ISAC framework for multiuser communication and target sensing. Specifically, we derive the Fisher information matrix (FIM) and the corresponding Cramér-Rao bound (CRB) for the joint estimation of target range, zenith angle, azimuth angle, and radial velocity. Based on the derived CRB, we jointly optimize the FAA port positions and transmit beamforming to improve sensing accuracy while satisfying communication signal-to-interference-plus-noise ratio (SINR), transmit power, and physical deployment constraints. To solve the resulting non-convex problem, we develop a two-stage alternating optimization algorithm, in which transmit beamforming is optimized through semidefinite relaxation (SDR), while FAA port positions are successively optimized through discrete search and continuous refinement. Simulation results indicate that the proposed algorithm achieves favorable sensing performance while maintaining a reasonable computational cost.
Conventional recommender systems rely on fixed cascaded pipelines, limiting their ability to adapt to evolving user needs and feedback. Existing agentic recommenders introduce reasoning and tool use but often lack end-to-end decision control and systematic failure reflection. We propose EvoRec, a unified agentic recommendation framework that combines autonomous decision-making with reflection-driven self-evolution. A Need Agent infers user needs from long-term, short-term, and working memory. A Planning Agent adaptively allocates retrieval budgets between Behavior Continuation and Need Satisfaction, selects retrieval tools, and sets ranking-fusion weights. An Acting Executor executes the plan and records decision traces. Following user feedback, a Dynamic Reflection Tree identifies failures in need inference, planning, or previous repairs, proposes personalized corrections, and validates them through replay. Verified repairs update working memory and are distilled into reusable cross-user strategies, enabling continual improvement without parameter updates. Experiments on a Kuaishou local-services advertising dataset show a 15.7% Hit@50 improvement over the strongest baseline. Hierarchical reflection further improves Hit@50 by 13.1% among users with prior repairs. In a seven-day online A/B test, nearline EvoRec increases advertising revenue by 4.6% and CTCVR by 3.5%, with a 14.5% revenue gain for high-intent users. These results demonstrate the effectiveness of reflection-driven self-evolution in industrial-scale recommendation.
Large language models (LLMs) have shown strong performance on software engineering (SE) tasks, driving the need for high-quality datasets for reliable training and evaluation. SWE-bench-style datasets (SWE datasets) are widely used to evaluate LLM-based coding agents on real-world software issues, where task clarity and test coverage directly affect whether instances are solvable and verifiable. However, quality issues persist in existing SWE datasets, and manual annotation is costly and labor-intensive, motivating LLM-based annotation. Existing approaches typically rely on static, overly general rubrics, limiting their ability to handle diverse and nuanced scenarios and often yielding high false positive rates. In contrast, human annotators draw on accumulated experience to make context-aware judgments. Inspired by this, we propose ExLabel, an LLM-based, self-evolving, experience-guided assistant for task clarity and test coverage assessment in SWE datasets. ExLabel maintains an evolving experience bank that distills cross-instance knowledge into stable yet adaptive judgment criteria and reuses relevant experience to guide the assessment of new instances. For systematic evaluation, we construct SWE-TQ, a new benchmark spanning eight programming languages, by cleaning human annotations from SWE-bench and Multi-SWE-bench. Experiments with three backbone LLMs show that ExLabel achieves the best F1 across all settings, with relative improvements of 44.2%/49.9% and 24.0%/18.4% over the best baseline for task clarity and test coverage assessment on the Python/Multilingual subsets, respectively. Moreover, a coding model fine-tuned on SWE-rebench data filtered by ExLabel resolves 66.2% of issues on SWE-bench Verified, outperforming the model trained on data filtered by the best baseline by 14.5%.
Jiayi Li, Guancheng Zhou, Yisi Luo, Ruihua Chen, Penghao Jiang, Deyu Meng, Flora D. Salim, Piotr Koniusz
Large-scale PDE pre-training trains a single model on the data from multiple partial differential equations (PDEs). However, adding heterogeneous PDEs with different governing equations, physical variables, and variable interactions does not always improve predictive performance across tasks. Existing PDE mixture-of-experts models mitigate cross-task interference through learned routing, but typically use homogeneous expert architectures, making it difficult to directly reflect differences in state evolution across equations. We introduce HotPOT, an operator transformer with heterogeneous physics experts for large-scale PDE pre-training. HotPOT uses a shared encoder to capture information common across PDEs, while each PDE family uses a neural-numerical expert that pairs a learnable predictor with a fixed numerical update containing no trainable parameters. Each predictor outputs the quantities required by its paired numerical update, which then computes the next state. This design allows different PDEs to share a common encoder while retaining their own learnable predictors and numerical updates. Experiments across five PDE families and six settings show that HotPOT outperforms MoE-POT in all 24 matched-scale comparisons before and after fine-tuning, reducing rollout error by 50.7% on average. Notably, the 17M model also surpasses substantially larger MoE-POT models, and the performance gains extend to zero-shot prediction under unseen physical conditions. Code is available at https://github.com/cruiseresearchgroup/HotPOT-code.
Jiayi Li, Penghao Jiang, Baiyu Chen, Yifan Duan, Hira Saleem, Zhaonan Wang, Piotr Koniusz, Flora D. Salim
Neural operators efficiently simulate dynamical systems by repeatedly predicting the next state. However, small errors at each step can accumulate, causing long-term predictions to drift from the true trajectory. Long-term error reflects both the propagation of earlier prediction errors and new errors introduced at each step. How to reduce these errors is therefore a critical question. We propose the Spectral Generator Neural Operator (SGNO), which combines a linear estimate of the next state with a learned correction. The linear component uses consecutive states in the training data to fit a map from the current state to the next. A neural network corrects this prediction by learning changes that the linear model does not capture, including nonlinear interactions. Both components are trained jointly toward the same next-state target. Experiments across 13 periodic PDE configurations in one to three dimensions show that SGNO accurately learns state evolution on linear tasks and reduces 100-step rollout error by 13-74% over the strongest baseline under the main comparison protocol on each nonlinear task. Longer-horizon extrapolation tests show lower late-window prediction errors on several nonlinear tasks. These results show that combining fitted propagation with learned state-dependent correction can improve long-horizon autoregressive prediction on the evaluated PDE tasks. Code is available at https://github.com/cruiseresearchgroup/SGNO.
Chenxing Wei, Jiazhen Kang, Hong Wang, Jianqing Zhang, Hao Jiang, Xiaolong Xu, Ningyuan Sun, Ying He 等 11 位作者
Reinforcement Learning with Verifiable Rewards (RLVR) successfully aligns autoregressive models in rigorous domains like mathematical reasoning and code generation. However, applying RLVR to Diffusion Large Language Models (dLLMs) is hindered by intractable sequence-level likelihoods, forcing existing methods to rely on unstable, high-variance approximations. We propose Sequence Likelihood-Free Policy Optimization (LFPO), an efficient framework that circumvents this bottleneck by formulating policy alignment as discrete velocity alignment. Inspired by continuous flow matching, LFPO eschews trajectory-level estimation and directly optimizes denoising logits via contrastive token-level updates, providing a stable, low-variance surrogate. Furthermore, predicting terminal solutions from intermediate states implicitly straightens the generation trajectory, enabling high-quality sampling with significantly fewer denoising iterations. Extensive experiments on reasoning and coding benchmarks demonstrate that LFPO matches or exceeds the accuracy of likelihood-based baselines while substantially accelerating inference.
Wenxuan Xie, Yaxun Dai Soochow University, Wenhao Jiang Guangming Laboratory
Recent advances in large language models (LLMs) have led to substantial progress on the Text-to-SQL task. However, existing approaches typically depend on static, pre-processed database information supplied at inference time, which restricts the model’s capacity to deeply comprehend the underlying database content. In the absence of dynamic interaction, LLMs are limited to fixed, human-curated context and lack the ability to autonomously query or explore the data. To overcome this limitation, we introduce \textbf{SDE-SQL}, a novel framework that empowers LLMs to perform \textbf{Self-Driven Exploration} of databases during inference. This is achieved through the generation and execution of \textbf{SQL probes}, enabling the model to actively retrieve information and iteratively refine its understanding of the database. Unlike prior methods, \textbf{SDE-SQL} operates in a \textbf{zero-shot} setting, requiring no in-context demonstrations or question-SQL pairs. Evaluated on the BIRD benchmark with Qwen2.5-72B-Instruct, \textbf{SDE-SQL} achieves an \textbf{8.02} % relative improvement in execution accuracy over the vanilla Qwen2.5-72B-Instruct baseline, establishing a new state-of-the-art among open-source methods without supervised fine-tuning (SFT) or model ensembling. Furthermore, when combined with SFT, \textbf{SDE-SQL} delivers an additional \textbf{0.52} % performance gain.
尚无 AI 解读，展示原文摘要
cs.CL
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表
Changhao Jiang, Ming Zhang Fudan University, Yifei Cao Fudan University, Junjie Ye, Xiaoran Fan, Shihan Dou, Zhiheng Xi, Jiajun Sun Fudan University 等 15 位作者
The GPT-4 technical report suggests that downstream performance can be predicted from pre-training signals, but offers little methodological detail on how to quantify this. This work address this gap by modeling knowledge retention, the capacity of a pre-trained language model to memorize factual information from its corpus, and introduce a principled method to estimate it prior to training. We propose Size-dependent Mutual Information (SMI), an information-theoretic predictor that integrates knowledge frequency, knowledge specificity, and model size to forecast closed-book question answering (QA) accuracy. SMI is validated through large-scale document retrieval over the disclosed pre-training corpora of 21 public and 3 custom models, combined with a robust multi-template QA evaluation. Experiments show that SMI significantly outperforms repetition-based baselines and achieves R² > 0.7 in predicting QA accuracy for models above 1B parameters, without additional training. The analysis further reveals diminishing returns from scaling data and model size and provides evidence for an intrinsic upper bound on knowledge retention achievable by pre-training alone, motivating retrieval and other augmentation strategies.
尚无 AI 解读，展示原文摘要
cs.CL
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表
Lei Wei, Xiao Peng Alibaba Group, Tt, Guannan Zhang, Chenhao Jiang, Hongyu Li Alibaba Group, Lanbo Lin, Yuanwu Xu Alibaba Group 等 11 位作者
Large Language Model (LLM) agents struggle with ultra-long-horizon tasks requiring hundreds or thousands of interaction steps. Traditional context management approaches face a fundamental dilemma: preserving complete histories rapidly exhausts context windows and forces crude truncation, while aggressive summarization discards critical information prematurely. We propose Predictive Adaptive Context Extraction (PACE), a novel framework that reconceptualizes context management as a Next Step Prediction problem. Inspired by neural attention, PACE dynamically constructs context by adjusting historical memory granularity based on its predicted relevance for the next action. Comprehensive evaluation across diverse benchmarks and models demonstrates that PACE consistently improves task success rates, with larger gains on complex tasks and robust cross-lingual performance. Crucially, PACE enables agents to sustain effective reasoning for 4,897 interaction steps in ultra-long-horizon scenarios, achieving a 66.2 improvement over the full-context ReAct baseline and 5.1 over advanced folding baselines. This fundamentally advances the capability of LLM-based agents in previously intractable long-horizon scenarios. Our code and data are available at https://anonymous.4open.science/r/PACE-B000/.
尚无 AI 解读，展示原文摘要
cs.CL
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表
Tao Feng Zhejiang University, Xinke Jiang, Xinyan Hu, Yonggang Zhang, Zhen Tao Harbin Institute of Technology, Shenzhen, Wentao Zhang, Boyang Liu, Wenhao Jiang Harbin Institute of Technology 等 9 位作者
Agentic search augments large language models (LLMs) with external knowledge through reinforcement learning. However, existing approaches suffer from blind reliance on noisy retrieval and hallucination when both parametric and external knowledge fail—reflecting a lack of calibration regarding the model’s knowledge boundary. We propose Knowledge boundary Policy Optimization (KbPO), a reinforcement learning framework that explicitly aligns retrieval decisions with quantified knowledge states. KbPO introduces: (1) a semantic stability metric to delineate reliable parametric knowledge; (2) a four-quadrant taxonomy synthesising internal certainty with retrieval quality; and (3) a quadrant-based reward mechanism incentivising calibrated behaviour. We further adopt an iterative query evolution pipeline to construct boundary-probing training samples. Experiments on ten benchmarks demonstrate that KbPO outperforms strong baselines while exhibiting reduced hallucination rates.
尚无 AI 解读，展示原文摘要
cs.CL · cs.AI
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表
Yaxun Dai Soochow University, Wenxuan Xie, Xialie Zhuang, Tianyu Yang Meituan, Ziyi Liu Soochow University, Haiqin Yang, Yiying Yang, Yuhang Zhao 等 10 位作者
Current Text-to-SQL reasoning models often lack integrated execution feedback during generation, and most existing approaches utilize feedback only for post-hoc correction. This separation not only limits real-time error correction, but may also introduce mistakes by altering otherwise correct SQL queries. To address these challenges, we present ReEx-SQL (Reasoning with Execution-Aware Reinforcement Learning), a Text-to-SQL framework that interacts with the SQL execution engine during decoding and dynamically adjusts reasoning based on execution feedback, thereby enabling context-sensitive query refinement and improved accuracy. ReEx-SQL achieves this through structured prompts with markup tags and a stepwise rollout strategy that incorporates execution feedback at each generation stage. For policy supervision, we design a composite reward function—featuring an exploration reward—to explicitly encourage effective interaction with the database. Furthermore, ReEx-SQL adopts a tree-based decoding strategy to facilitate exploratory reasoning and primarily aims to enhance parallel decoding efficiency. Notably, ReEx-SQL achieves 89.1% accuracy on Spider and 65.3% on BIRD at the 7B scale, surpassing baseline models by 2.7% and 2.6%, respectively. In addition, its tree-based decoding accelerates inference by 51.9% compared to linear decoding during sampling.
尚无 AI 解读，展示原文摘要
cs.CL
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表
Recent progress in large language models (LLMs) is largely driven by scaling training compute through either pre-training with next-token prediction (NTP) or post-training with reinforcement learning (RL). The former contributes to learning broad knowledge and skills from general data, while struggling with data inefficiency and catastrophic forgetting in continual learning settings. The latter incentivizes reasoning capabilities with strong generalization, but is constrained by limited data availability due to its reliance on human annotation. To alleviate these issues, we propose Reinforcement Learning on Pre-Training data (RLPT), which combines the advantages of learning from general data and RL. In particular, RLPT derives reward signals directly from general text data through a next-segment reasoning objective, rewarding the policy for correctly predicting next text segments conditioned on the prefix text. Experiments across multiple benchmarks and models demonstrate the effectiveness of . For example, RLPT yields substantial improvements in continual pre-training (+4.6\%) and provides a strong foundation for post-training (+3.4\%) on Qwen3-8B-Base.
尚无 AI 解读，展示原文摘要
cs.CL
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表
Hongchao Jiang Nanyang Technological University, Yiming Chen, Yushi Cao AICS, Hung-yi Lee National Taiwan University, Robby T. Tan
Large Language Models (LLMs) are increasingly used not only to generate code, but also to judge it: comparing, ranking, or scoring competing solutions. However, their reliability in this evaluative role remains poorly understood. Inconsistent or flawed judgments can undermine benchmarks and distort training signals. This paper investigates the performance and robustness of LLMs when used as code judges. We introduce CodeJudgeBench, a benchmark explicitly designed to evaluate LLM-as-a-Judge models across three critical coding tasks: code generation, code repair, and unit test generation. We comprehensively benchmark the performance of 26 LLM-as-a-Judge models, encompassing general-purpose, code-tuned, and reasoning models. Our empirical findings reveal that relatively small reasoning models (e.g., Qwen3-8B) can outperform much larger non-reasoning models up to 70B. We further stress-test robustness by applying both general and code-specific perturbations. All models show significant instability and are sensitive to changes such as response ordering, variable naming, and misleading comments. These findings highlight serious concerns about the consistency and robustness of LLM-based judges for coding tasks.
尚无 AI 解读，展示原文摘要
cs.CL
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表
Ming Zhang Fudan University, Yujiong Shen, Jingyi Deng, Yuhui Wang, Huayu Sha, Kexin Tan The University of Tokyo, The University of Tokyo and Fudan University, Qiyuan Peng, Yue Zhang ATOM INFINITE PTE. LTD. 等 22 位作者
Existing evaluation of Large Language Models (LLMs) on static benchmarks is vulnerable to data contamination and leaderboard overfitting, critical issues that obscure true model capabilities. To address this, we introduce LLMEval-Fair, a framework for dynamic evaluation of LLMs. LLMEval-Fair is built on a proprietary bank of 220k graduate-level questions, from which it dynamically samples unseen test sets for each evaluation run. Its automated pipeline ensures integrity via contamination-resistant data curation, a novel anti-cheating architecture, and a calibrated LLM-as-a-judge process achieving 90% agreement with human experts, complemented by a relative ranking system for fair comparison. An 30-month longitudinal study of nearly 60 leading models reveals a performance ceiling on knowledge memorization and exposes data contamination vulnerabilities undetectable by static benchmarks. The framework demonstrates exceptional robustness in ranking stability and consistency, providing strong empirical validation for the dynamic evaluation paradigm. LLMEval-Fair offers a robust and credible methodology for assessing the true capabilities of LLMs beyond leaderboard scores, promoting the development of more trustworthy evaluation standards.
尚无 AI 解读，展示原文摘要
cs.CL
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表