AI Agent - arXivDaily 专题

2512.09679 2025-12-11 cs.SE 57%

Understanding Chain-of-Thought Effectiveness in Code Generation: An Empirical and Information-Theoretic Analysis

理解链式思维在代码生成中的有效性：一项经验性与信息论分析

Naizhu Jin, Zhong Li, Guang Yang, Tian Zhang, Qingkai Zeng

专题命中软件智能体：planning(abstract)；分类 cs.SE

AI总结本研究通过经验分析和信息论方法，揭示了链式思维在代码生成中有效性的影响因素，发现结构化CoT在提升生成质量方面表现更优，且其效果受语言系统和模型容量影响。

2512.07302 2025-12-09 cs.CV cs.AI 57%

Towards Accurate UAV Image Perception: Guiding Vision-Language Models with Stronger Task Prompts

迈向准确的无人机图像感知：通过更强的任务提示引导视觉-语言模型

Mingning Guo, Mengwei Wu, Shaoxian Li, Haifeng Li, Chao Tao

机构 * School of Geosciences and InfoPhysics, Central South University（地质科学与信息物理学院，中南大学）

专题命中软件智能体：agent(abstract)；分类 cs.AI

AI总结 AerialVP通过增强任务提示提升无人机图像感知性能，引入AerialSense基准评估模型在复杂场景下的表现。

2506.23749 2025-12-05 cs.SE 57%

A Survey of LLM-based Automated Program Repair: Taxonomies, Design Paradigms, and Applications

基于大型语言模型的自动程序修复综述：分类、设计范式与应用

Boyang Yang, Zijian Cai, Fengling Liu, Bach Le, Lingming Zhang, Tegawendé F. Bissyandé, Yang Liu, Haoye Tian

专题命中软件智能体：agentic(abstract)；分类 cs.SE

AI总结本文综述了基于LLM的自动程序修复，提出统一分类法，涵盖四种范式，并讨论了评估实践和未来研究方向。

2511.23302 2025-12-01 cs.SE 57%

FLIMs: Fault Localization Interference Mutants, Definition, Recognition and Mitigation

FLIMs: 故障定位干扰突变体，定义、识别与缓解

Hengyuan Liu, Zheng Li, Donghua Wang, Yankai Wu, Xiang Chen, Yong Liu

专题命中软件智能体：workflow(abstract)；分类 cs.SE

AI总结本文提出MBFL-FLIM框架，通过识别和缓解干扰突变体提升故障定位效果，实验证明其在多故障场景中表现优异。

2511.03187 2025-12-01 cs.LG cs.RO 57%

Periodic Skill Discovery

周期性技能发现

Jonghae Park, Daesol Cho, Jusuk Lee, Dongseok Shim, Inkyu Jang, H. Jin Kim

专题命中软件智能体：agent(abstract)；分类 cs.LG

AI总结周期性技能发现（PSD）通过无监督学习发现具有不同周期的技能，提升复杂机器人任务中的表现和多样性。

Comments NeurIPS 2025

2511.14169 2025-11-25 cs.CV cs.AI 57%

AdaTok: Adaptive Token Compression with Object-Aware Representations for Efficient Multimodal LLMs

AdaTok: 一种基于对象感知表示的自适应令牌压缩方法，用于高效多模态大语言模型

Xinliang Zhang, Lei Zhu, Hangzhou He, Shuang Zeng, Ourui Fu, Jiakui Hu, Zhengjian Yao, Yanye Lu

机构 * Institute of Medical Technology, Peking University Health Science Center（北京大学医学部医学技术研究所）； Department of Biomedical Engineering, Peking University（北京大学生物医学工程系）； National Biomedical Imaging Center, Peking University（北京大学国家生物医学成像中心）

专题命中软件智能体：workflow(abstract)；分类 cs.AI

AI总结 AdaTok通过自适应令牌压缩方法，利用对象感知表示提升多模态大语言模型的效率，实现高压缩比与高性能的平衡。

2405.18300 2025-11-20 cs.AI 57%

CompetEvo: Towards Morphological Evolution from Competition

Kangyao Huang, Di Guo, Xinyu Zhang, Xiangyang Ji, Huaping Liu

机构 * Department of Computer Science and Technology, Tsinghua University（清华大学计算机科学与技术系）； School of Artificial Intelligence, Beijing University of Posts and Telecommunications（北京邮电大学人工智能学院）； School of Vehicle and Mobility, Tsinghua University（清华大学车辆与移动性学院）； Department of Automation, Tsinghua University（清华大学自动化系）

专题命中软件智能体：agent(abstract)；分类 cs.AI

2502.15980 2025-11-17 cs.HC cs.AI cs.DB 57%

Text-to-SQL Domain Adaptation via Human-LLM Collaborative Data Annotation

Yuan Tian, Daniel Lee, Fei Wu, Tung Mai, Kun Qian, Siddhartha Sahai, Tianyi Zhang, Yunyao Li

机构 * Purdue University（普渡大学）； Adobe Inc.（Adobe公司）

专题命中软件智能体：workflow(abstract)；分类 cs.AI

Comments Accepted by IUI'25 Code & Demo: https://github.com/magic-YuanTian/SQLsynth

2508.03768 2025-11-12 cs.LG 57%

ORVIT: Near-Optimal Online Distributionally Robust Reinforcement Learning

Debamita Ghosh, George K. Atia, Yue Wang

机构 * Department of Electrical and Computer Engineering（电气与计算机工程系）； Department of Computer Science（计算机科学系）； University of Central Florida（中央佛罗里达大学）

专题命中软件智能体：agent(abstract)；分类 cs.LG

Comments Accepted by AAAI 2026

2407.19405 2025-11-11 cs.AI 57%

Logic Distillation: Learning from Code Function by Function for Decision-making Tasks

Dong Chen, Shilin Zhang, Fei Gao, Yueting Zhuang, Siliang Tang, Qidong Liu, Mingliang Xu

机构 * The School of Computer and Artificial Intelligence of Zhengzhou University（郑州大学计算机与人工智能学院）； Engineering Research Center of Intelligent Swarm Systems, Ministry of Education（教育部智能群体系统工程研究中心）； National Supercomputing Center In Zhengzhou（郑州国家超级计算中心）； Zhejiang University（浙江大学）

专题命中软件智能体：planning(abstract)；分类 cs.AI

Comments 9 pages, 7 figures

2511.03586 2025-11-06 cs.PF cs.AI 57%

PerfDojo: Automated ML Library Generation for Heterogeneous Architectures

Andrei Ivanov, Siyuan Shen, Gioele Gottardo, Marcin Chrapek, Afif Boudaoud, Timo Schneider, Luca Benini, Torsten Hoefler

机构 * ETH Zurich（苏黎世联邦理工学院）

专题命中软件智能体：agent(abstract)；分类 cs.AI

Journal ref The International Conference for High Performance Computing, Networking, Storage and Analysis (SC '25), November 16--21, 2025, St Louis, MO, USA

2511.01183 2025-11-05 cs.AI cs.PL 57%

QiMeng-NeuComBack: Self-Evolving Translation from IR to Assembly Code

Hainan Fang, Yuanbo Wen, Jun Bi, Yihan Wang, Tonghui He, Yanlin Tang, Di Huang, Jiaming Guo, Rui Zhang, Qi Guo, Yunji Chen

机构 * State Key Lab of Processors, Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China（中国科学院计算技术研究所处理器重点实验室，北京，中国）； University of Chinese Academy of Sciences, Beijing, China（中国科学院大学，北京，中国）

专题命中软件智能体：workflow(abstract)；分类 cs.AI

Comments Accepted at NeurIPS 2025

2506.09485 2025-11-05 cs.RO cs.AI cs.GR 57%

Adv-BMT: Bidirectional Motion Transformer for Safety-Critical Traffic Scenario Generation

Yuxin Liu, Zhenghao Peng, Xuanhao Cui, Bolei Zhou

机构 * University of California, Los Angeles（加州大学洛杉矶分校）

专题命中软件智能体：agent(abstract)；分类 cs.AI

2511.00041 2025-11-04 cs.RO cs.AI 57%

Endowing GPT-4 with a Humanoid Body: Building the Bridge Between Off-the-Shelf VLMs and the Physical World

Yingzhao Jian, Zhongan Wang, Yi Yang, Hehe Fan

机构 * College of Computer Science and Technology（计算机科学与技术学院）； Zhejiang University（浙江大学）

专题命中软件智能体：agent(abstract)；分类 cs.AI

2510.25406 2025-10-31 cs.SE 57%

Dissect-and-Restore: AI-based Code Verification with Transient Refactoring

Changjie Wang, Mariano Scazzariello, Anoud Alshnakat, Roberto Guanciale, Dejan Kostić, Marco Chiesa

专题命中软件智能体：workflow(abstract)；分类 cs.SE

2510.22530 2025-10-28 cs.SE 57%

Finding the Needle in the Crash Stack: Industrial-Scale Crash Root Cause Localization with AutoCrashFL

Sungmin Kang, Sumi Yun, Jingun Hong, Shin Yoo, Gabin An

专题命中软件智能体：agent(abstract)；分类 cs.SE

Comments 11 pages, 8 figures, under review

2510.19002 2025-10-23 cs.GT cs.LG econ.TH math.OC 57%

Impartial Selection with Predictions

Javier Cembrano, Felix Fischer, Max Klimm

专题命中软件智能体：agent(abstract)；分类 cs.LG

2510.18491 2025-10-22 cs.AI 57%

Crucible: Quantifying the Potential of Control Algorithms through LLM Agents

Lianchen Jia, Chaoyang Li, Qian Houde, Tianchi Huang, Jiangchuan Liu, Lifeng Sun

机构 * Department of Computer Science and Technology, Tsinghua University（清华大学计算机科学与技术系）； Simon Fraser University（西蒙 Fraser大学）

专题命中软件智能体：agent(abstract)；分类 cs.AI

Comments NeurIPS 2025

2509.06239 2025-10-22 cs.AI 57%

Proof2Silicon: Prompt Repair for Verified Code and Hardware Generation via Reinforcement Learning

Manvi Jha, Jiaxin Wan, Deming Chen

机构 * Electrical and Computer Engineering（电气与计算机工程系）； University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）

专题命中软件智能体：agent(abstract)；分类 cs.AI

2510.17862 2025-10-22 cs.CR cs.SE 57%

When "Correct" Is Not Safe: Can We Trust Functionally Correct Patches Generated by Code Agents?

Yibo Peng, James Song, Lei Li, Xinyu Yang, Mihai Christodorescu, Ravi Mangal, Corina Pasareanu, Haizhong Zheng, Beidi Chen

专题命中软件智能体：agent(abstract)；分类 cs.SE

2510.15906 2025-10-21 cs.AR cs.AI 57%

FVDebug: An LLM-Driven Debugging Assistant for Automated Root Cause Analysis of Formal Verification Failures

Yunsheng Bai, Ghaith Bany Hamad, Chia-Tung Ho, Syed Suhaib, Haoxing Ren

机构 * NVIDIA

专题命中软件智能体：agentic(abstract)；分类 cs.AI

2510.15898 2025-10-21 cs.HC cs.CL cs.CY 57%

HealthDial: A No-Code LLM-Assisted Dialogue Authoring Tool for Healthcare Virtual Agents

Farnaz Nouraei, Zhuorui Yong, Timothy Bickmore

机构 * Northeastern University（东北大学）

专题命中软件智能体：agent(abstract)；分类 cs.CL

2510.12428 2025-10-15 cs.AI 57%

Biased-Attention Guided Risk Prediction for Safe Decision-Making at Unsignalized Intersections

Chengyang Dong, Nan Guo

专题命中软件智能体：agent(abstract)；分类 cs.AI

2510.11151 2025-10-14 cs.CL cs.CR 57%

TypePilot: Leveraging the Scala Type System for Secure LLM-generated Code

Alexander Sternfeld, Andrei Kucharavy, Ljiljana Dolamic

机构 * Institute of Entrepreneurship & Management, HES-SO Le Foyer, Techno-Pôle 1 Sierre, Switzerland（创业与管理学院，HES-SO莱福院，技术园区1，瑞士）； Institute of Informatics, HES-SO Techno-Pôle 3 Sierre, Switzerland（信息学院，HES-SO技术园区3，瑞士）； Cyber-Defence Campus armasuisse, Science and Technology Thun, Switzerland（网络安全校区，armasuisse，科技，瑞士）

专题命中软件智能体：agentic(abstract)；分类 cs.CL

2510.10551 2025-10-14 cs.SE 57%

How Students Use Generative AI for Software Testing: An Observational Study

Baris Ardic, Quentin Le Dilavrec, Andy Zaidman

专题命中软件智能体：workflow(abstract)；分类 cs.SE

Comments 39 pages, 6 figures, journal submission

2510.06056 2025-10-08 cs.AI 57%

Scientific Algorithm Discovery by Augmenting AlphaEvolve with Deep Research

Gang Liu, Yihan Zhu, Jie Chen, Meng Jiang

机构 * University of Notre Dame（诺丁汉大学）； MIT-IBM Watson AI Lab, IBM Research（麻省理工-IBM Watson AI实验室，IBM研究）

专题命中软件智能体：agent(abstract)；分类 cs.AI

Comments 25 pages, 17 figures, 4 tables

2510.05327 2025-10-08 cs.AR cs.AI 57%

DeepV: A Model-Agnostic Retrieval-Augmented Framework for Verilog Code Generation with a High-Quality Knowledge Base

Zahin Ibnat, Paul E. Calzada, Rasin Mohammed Ihtemam, Sujan Kumar Saha, Jingbo Zhou, Farimah Farahmandi, Mark Tehranipoor

机构 * University of Florida（佛罗里达大学）

专题命中软件智能体：agentic(abstract)；分类 cs.AI

Comments 22 pages, 6 figures

2510.04073 2025-10-07 cs.AI 57%

Moral Anchor System: A Predictive Framework for AI Value Alignment and Drift Prevention

Santhosh Kumar Ravindran

机构 * Microsoft Corporation（微软公司）

专题命中软件智能体：AI agent(abstract)；分类 cs.AI

Comments 11 pages Includes simulations with over 4 million steps

2509.21427 2025-10-07 cs.SE 57%

Extracting Conceptual Knowledge to Locate Software Issues

Ying Wang, Wenjun Mao, Chong Wang, Zhenhao Zhou, Yicheng Zhou, Wenyun Zhao, Yiling Lou, Xin Peng

专题命中软件智能体：agent(abstract)；分类 cs.SE

2509.25894 2025-10-01 cs.SE 57%

Red Teaming Program Repair Agents: When Correct Patches can Hide Vulnerabilities

Simin Chen, Yixin He, Suman Jana, Baishakhi Ray

专题命中软件智能体：agent(abstract)；分类 cs.SE

1. 软件智能体 3270 篇

Understanding Chain-of-Thought Effectiveness in Code Generation: An Empirical and Information-Theoretic Analysis

Towards Accurate UAV Image Perception: Guiding Vision-Language Models with Stronger Task Prompts

A Survey of LLM-based Automated Program Repair: Taxonomies, Design Paradigms, and Applications

FLIMs: Fault Localization Interference Mutants, Definition, Recognition and Mitigation

Periodic Skill Discovery

AdaTok: Adaptive Token Compression with Object-Aware Representations for Efficient Multimodal LLMs

CompetEvo: Towards Morphological Evolution from Competition

Text-to-SQL Domain Adaptation via Human-LLM Collaborative Data Annotation

ORVIT: Near-Optimal Online Distributionally Robust Reinforcement Learning

Logic Distillation: Learning from Code Function by Function for Decision-making Tasks

PerfDojo: Automated ML Library Generation for Heterogeneous Architectures

QiMeng-NeuComBack: Self-Evolving Translation from IR to Assembly Code

Adv-BMT: Bidirectional Motion Transformer for Safety-Critical Traffic Scenario Generation

Endowing GPT-4 with a Humanoid Body: Building the Bridge Between Off-the-Shelf VLMs and the Physical World

Dissect-and-Restore: AI-based Code Verification with Transient Refactoring

Finding the Needle in the Crash Stack: Industrial-Scale Crash Root Cause Localization with AutoCrashFL

Impartial Selection with Predictions

Crucible: Quantifying the Potential of Control Algorithms through LLM Agents

Proof2Silicon: Prompt Repair for Verified Code and Hardware Generation via Reinforcement Learning

When "Correct" Is Not Safe: Can We Trust Functionally Correct Patches Generated by Code Agents?

FVDebug: An LLM-Driven Debugging Assistant for Automated Root Cause Analysis of Formal Verification Failures

HealthDial: A No-Code LLM-Assisted Dialogue Authoring Tool for Healthcare Virtual Agents

Biased-Attention Guided Risk Prediction for Safe Decision-Making at Unsignalized Intersections

TypePilot: Leveraging the Scala Type System for Secure LLM-generated Code

How Students Use Generative AI for Software Testing: An Observational Study

Scientific Algorithm Discovery by Augmenting AlphaEvolve with Deep Research

DeepV: A Model-Agnostic Retrieval-Augmented Framework for Verilog Code Generation with a High-Quality Knowledge Base

Moral Anchor System: A Predictive Framework for AI Value Alignment and Drift Prevention

Extracting Conceptual Knowledge to Locate Software Issues

Red Teaming Program Repair Agents: When Correct Patches can Hide Vulnerabilities