大模型推理能力

2506.12713 2025-10-21 cs.SE cs.CL 57%

Humanity's Last Code Exam: Can Advanced LLMs Conquer Human's Hardest Code Competition?

Xiangyang Li, Xiaopeng Li, Kuicai Dong, Quanhu Zhang, Rongju Ruan, Xinyi Dai, Xiaoshuang Liu, Shengchun Xu, Yasheng Wang, Ruiming Tang

机构 * Huawei Noah’s Ark Lab（华为诺亚实验室）

专题命中推理评测：reasoning(abstract)；分类 cs.CL

Comments EMNLP 2025 Findings

2507.05289 2025-10-21 cs.SE 50%

Measuring how changes in code readability attributes affect code quality evaluation by Large Language Models

Igor Regis da Silva Simoes, Elaine Venson

专题命中推理评测：reasoning(abstract)

Journal ref 2025: Proceedings of the XXXIX Brazilian Symposium on Software Engineering

2505.14671 2025-10-21 cs.CV 50%

UniCTokens: Boosting Personalized Understanding and Generation via Unified Concept Tokens

Ruichuan An, Sihan Yang, Renrui Zhang, Zijun Shen, Ming Lu, Gaole Dai, Hao Liang, Ziyu Guo, Shilin Yan, Yulin Luo, Bocheng Zou, Chaoqun Yang, Wentao Zhang

机构 * Peking University（北京大学）； Xi’an JiaoTong University（西安交通大学）； CUHK（香港中文大学）； Intel Labs, China（中国英特尔实验室）； Nanjing University（南京大学）； University of Wisconsin-Madison（威斯康星大学麦迪逊分校）； Tsinghua University（清华大学）

专题命中推理评测：reasoning(abstract)

Journal ref NeurIPS 2025

2510.16641 2025-10-21 cs.CV 50%

MultiVerse: A Multi-Turn Conversation Benchmark for Evaluating Large Vision and Language Models

Young-Jun Lee, Byung-Kwan Lee, Jianshu Zhang, Yechan Hwang, Byungsoo Ko, Han-Gyu Kim, Dongyu Yao, Xuankun Rong, Eojin Joo, Seung-Ho Han, Bowon Ko, Ho-Jin Choi

机构 * KAIST（韩国科学技术院）； WHU（华中科技大学）； NAVER ； CMU（卡内基梅隆大学）

专题命中推理评测：reasoning(abstract)

Comments Project website: https://passing2961.github.io/multiverse-project-page/

2510.16579 2025-10-21 cs.SE 50%

Human-Aligned Code Readability Assessment with Large Language Models

Wendkûuni C. Ouédraogo, Yinghua Li, Xueqi Dang, Pawel Borsukiewicz, Xin Zhou, Anil Koyuncu, Jacques Klein, David Lo, Tegawendé F. Bissyandé

专题命中推理评测：CoT(abstract)

2510.16137 2025-10-21 physics.ed-ph 50%

Exploring the potential of ChatGPT for feedback and evaluation in experimental physics

Marcos Abreu, Álvaro Suárez, Cecilia Stari, Arturo C. Marti

专题命中推理评测：reasoning(abstract)

Comments 5 pages

2508.02736 2025-10-21 cs.OS cs.SE 50%

AgentSight: System-Level Observability for AI Agents Using eBPF

Yusheng Zheng, Yanpeng Hu, Tong Yu, Andi Quinn

专题命中推理评测：reasoning(abstract)

Journal ref PACMI'2025

2510.09211 2025-10-21 cs.CL cs.AI 90%

DICE: Structured Reasoning in LLMs through SLM-Guided Chain-of-Thought Correction

Yiqi Li, Yusheng Liao, Zhe Chen, Yanfeng Wang, Yu Wang

机构 * Shanghai Jiao Tong University（上海交通大学）； Shanghai Artificial Intelligence Laboratory（上海人工智能实验室）

专题命中其他推理：reasoning(title,abstract)；chain-of-thought(title,abstract)；CoT(abstract)；分类 cs.CL、cs.AI

Comments This paper was accepted to the EMNLP 2025 main conference

2510.16761 2025-10-21 cs.CL 79%

Enhancing Language Agent Strategic Reasoning through Self-Play in Adversarial Games

Yikai Zhang, Ye Rong, Siyu Yuan, Jiangjie Chen, Jian Xie, Yanghua Xiao

专题命中其他推理：reasoning(title,abstract)；分类 cs.CL

2510.16455 2025-10-21 cs.CL 79%

RAVEN: Robust Advertisement Video Violation Temporal Grounding via Reinforcement Reasoning

Deyi Ji, Yuekui Yang, Haiyang Wu, Shaoping Ma, Tianrun Chen, Lanyun Zhu

机构 * Tencent（腾讯公司）； Department of Computer Science and Technology, Tsinghua University（清华大学计算机科学与技术系）； Zhejiang University（浙江大学）； Singapore University of Technology and Design（新加坡科技设计大学）

专题命中其他推理：reasoning(title,abstract)；分类 cs.CL

Comments ACL 2025 (Oral, Industry Track)

2510.15990 2025-10-21 cs.LG cs.AI cs.CL 67%

Can GRPO Help LLMs Transcend Their Pretraining Origin?

Kangqi Ni, Zhen Tan, Zijie Liu, Pingzhi Li, Tianlong Chen

机构 * cs.unc.edu（北卡罗来纳大学教堂山分校计算机科学系）

专题命中其他推理：reasoning(abstract)；分类 cs.CL、cs.AI、cs.LG

2505.16984 2025-10-21 cs.LG cs.CL 62%

UFT: Unifying Supervised and Reinforcement Fine-Tuning

Mingyang Liu, Gabriele Farina, Asuman Ozdaglar

专题命中其他推理：reasoning(abstract)；分类 cs.CL、cs.LG

2510.16276 2025-10-21 cs.AI cs.LG 62%

What Limits Agentic Systems Efficiency?

Song Bian, Minghao Yan, Anand Jayarajan, Gennady Pekhimenko, Shivaram Venkataraman

机构 * UW-Madison（威斯康星大学麦迪逊分校）； University of Toronto（多伦多大学）； NVIDIA（英伟达）

专题命中其他推理：reasoning(abstract)；分类 cs.AI、cs.LG

Comments 27 pages, 15 figures

2510.16933 2025-10-21 cs.DC cs.AI 57%

Tutoring LLM into a Better CUDA Optimizer

Matyáš Brabec, Jiří Klepl, Michal Töpfer, Martin Kruliš

机构 * Charles University（查理大学）

专题命中其他推理：reasoning(abstract)；分类 cs.AI

Comments This preprint has not undergone peer review or any post-submission improvements or corrections. The Version of Record of this contribution is published in Euro-Par 2025: Parallel Processing, Part II, and is available online at https://doi.org/10.1007/978-3-031-99857-7_18

Journal ref Euro-Par 2025: Parallel Processing. Euro-Par 2025. Lecture Notes in Computer Science, vol 15901. Springer, Cham

2510.16885 2025-10-21 cs.LG 57%

UniGTE: Unified Graph-Text Encoding for Zero-Shot Generalization across Graph Tasks and Domains

Duo Wang, Yuan Zuo, Guangyue Lu, Junjie Wu

机构 * MIIT Key Laboratory of Data Intelligence and Management, Beihang University（信息科技部数据智能与管理重点实验室，北京航空航天大学）

专题命中其他推理：reasoning(abstract)；分类 cs.LG

Journal ref NeurIPS 2025

2506.04650 2025-10-21 cs.LG 57%

Neural Network Reprogrammability: A Unified Theme on Model Reprogramming, Prompt Tuning, and Prompt Instruction

Zesheng Ye, Chengyi Cai, Ruijiang Dong, Jianzhong Qi, Lei Feng, Pin-Yu Chen, Feng Liu

专题命中其他推理：chain-of-thought(abstract)；分类 cs.LG

2510.16290 2025-10-21 cs.CV cs.CL 57%

Cerberus: Real-Time Video Anomaly Detection via Cascaded Vision-Language Models

Yue Zheng, Xiufang Shi, Jiming Chen, Yuanchao Shu

机构 * Zhejiang University of Technology（浙江工业大学）； Zhejiang University（浙江大学）； Hangzhou Dianzi University（杭州电子科技大学）

专题命中其他推理：reasoning(abstract)；分类 cs.CL

2510.16080 2025-10-21 q-bio.QM cs.AI 57%

TriAgent: Automated Biomarker Discovery with Deep Research Grounding for Triage in Acute Care by LLM-Based Multi-Agent Collaboration

Kerem Delikoyun, Qianyu Chen, Win Sen Kuan, John Tshon Yit Soong, Matthew Edward Cove, Oliver Hayden

机构 * Technical University of Munich（慕尼黑技术大学）； National University of Singapore（新加坡国立大学）； National University Hospital（新加坡国立大学医院）

专题命中其他推理：CoT(abstract)；分类 cs.AI

2510.15917 2025-10-21 cs.AR cs.AI cs.DC 57%

Intent-Driven Storage Systems: From Low-Level Tuning to High-Level Understanding

Shai Bergman, Won Wook Song, Lukas Cavigelli, Konstantin Berestizshevsky, Ke Zhou, Ji Zhang

机构 * Huawei Zurich Research Center（华为苏黎世研究中心）； Huazhong University of Science and Technology（华中科技大学）

专题命中其他推理：reasoning(abstract)；分类 cs.AI

2510.17777 2025-10-21 cs.CV 50%

SparseVILA: Decoupling Visual Sparsity for Efficient VLM Inference

Samir Khaki, Junxian Guo, Jiaming Tang, Shang Yang, Yukang Chen, Konstantinos N. Plataniotis, Yao Lu, Song Han, Zhijian Liu

机构 * NVIDIA ； MIT（麻省理工学院）； UC San Diego（南加州大学圣地亚哥分校）； University of Toronto（多伦多大学）

专题命中其他推理：reasoning(abstract)

2510.17164 2025-10-21 cs.SE 50%

Software Testing with Large Language Models: An Interview Study with Practitioners

Maria Deolinda Santana, Cleyton Magalhaes, Ronnie de Souza Santos

专题命中其他推理：reasoning(abstract)

2510.16418 2025-10-21 cs.DC 50%

FourierCompress: Layer-Aware Spectral Activation Compression for Efficient and Accurate Collaborative LLM Inference

Jian Ma, Xinchen Lyu, Jun Jiang, Longhao Zou, Chenshan Ren, Qimei Cui, Xiaofeng Tao

专题命中其他推理：reasoning(abstract)

2510.16366 2025-10-21 cs.CY 50%

Integrating LLM and Diffusion-Based Agents for Social Simulation

Xinyi Li, Zhiqiang Guo, Qinglang Guo, Hao Jin, Weizhi Ma, Min Zhang

专题命中其他推理：reasoning(abstract)

Comments 10 pages, 3 figures, 4 tables

1. 推理评测 42 篇

Humanity's Last Code Exam: Can Advanced LLMs Conquer Human's Hardest Code Competition?

Measuring how changes in code readability attributes affect code quality evaluation by Large Language Models

UniCTokens: Boosting Personalized Understanding and Generation via Unified Concept Tokens

MultiVerse: A Multi-Turn Conversation Benchmark for Evaluating Large Vision and Language Models

Human-Aligned Code Readability Assessment with Large Language Models

Exploring the potential of ChatGPT for feedback and evaluation in experimental physics

AgentSight: System-Level Observability for AI Agents Using eBPF

2. 其他推理 16 篇

DICE: Structured Reasoning in LLMs through SLM-Guided Chain-of-Thought Correction

Enhancing Language Agent Strategic Reasoning through Self-Play in Adversarial Games

RAVEN: Robust Advertisement Video Violation Temporal Grounding via Reinforcement Reasoning

Can GRPO Help LLMs Transcend Their Pretraining Origin?

UFT: Unifying Supervised and Reinforcement Fine-Tuning

What Limits Agentic Systems Efficiency?

Tutoring LLM into a Better CUDA Optimizer

UniGTE: Unified Graph-Text Encoding for Zero-Shot Generalization across Graph Tasks and Domains

Neural Network Reprogrammability: A Unified Theme on Model Reprogramming, Prompt Tuning, and Prompt Instruction

Cerberus: Real-Time Video Anomaly Detection via Cascaded Vision-Language Models

TriAgent: Automated Biomarker Discovery with Deep Research Grounding for Triage in Acute Care by LLM-Based Multi-Agent Collaboration

Intent-Driven Storage Systems: From Low-Level Tuning to High-Level Understanding

SparseVILA: Decoupling Visual Sparsity for Efficient VLM Inference

Software Testing with Large Language Models: An Interview Study with Practitioners

FourierCompress: Layer-Aware Spectral Activation Compression for Efficient and Accurate Collaborative LLM Inference

Integrating LLM and Diffusion-Based Agents for Social Simulation