大模型推理能力

A2R: An Asymmetric Two-Stage Reasoning Framework for Parallel Reasoning

Ziqi Wang, Boye Niu, Zhongli Li, Linghui Meng, Jing Liu, Zhi Zheng, Tong Xu, Hua Wu, Haifeng Wang, Enhong Chen

机构 * USTC（中国科学技术大学） ； Baidu（百度） ； USYD（悉尼大学） 纠错

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.CL、cs.AI

Comments 15 pages, 3 figures

CCQA: Generating Question from Solution Can Improve Inference-Time Reasoning in SLMs

Jin Young Kim, Ji Won Yoon

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.CL、cs.AI

Comments Published as a main conference paper at EMNLP 2025

SpecVLM: Enhancing Speculative Decoding of Video LLMs via Verifier-Guided Token Pruning

Yicheng Ji, Jun Zhang, Heming Xia, Jinpeng Chen, Lidan Shou, Gang Chen, Huan Li

机构 * The State Key Laboratory of Blockchain and Data Security（区块链与数据安全国家重点实验室） ； Hangzhou High-Tech Zone (Binjiang) Institute of Blockchain and Data Security（杭州高新技术区（滨江）区块链与数据安全研究院） ； Department of Computing, The Hong Kong Polytechnic University（香港理工大学计算机系） ； School of Computer Science, Beijing University of Posts and Telecommunications（北京邮电大学计算机学院） 纠错

专题命中 测试时计算 ：verifier(title,abstract)；分类 cs.CL、cs.AI

Comments Accepted at EMNLP 2025 Main

RTTC: Reward-Guided Collaborative Test-Time Compute

J. Pablo Muñoz, Jinjie Yuan

机构 * Intel Labs Santa Clara, CA, USA（英特尔圣克拉拉实验室） ； Intel Corporation Beijing, China（英特尔北京公司） 纠错

专题命中 测试时计算 ：test-time compute(title,abstract)；分类 cs.CL、cs.AI

ReasoningGuard: Safeguarding Large Reasoning Models with Inference-time Safety Aha Moments

Yuquan Wang, Mi Zhang, Yining Wang, Geng Hong, Xiaoyu You, Min Yang

机构 * Fudan University（复旦大学） ； Shanghai University of Electric Power（上海电力大学） ； East China University of Science and Technology（东华大学） 纠错

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.CL、cs.AI

SmartThinker: Learning to Compress and Preserve Reasoning by Step-Level Length Control

Xingyang He, Xiao Ling, Jie Liu

机构 * Xingyang He 1（He Xingyang 1） ； Xiao Ling 1（Ling Xiao 1） ； Jie Liu 1（Liu Jie 1） 纠错

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.CL、cs.AI

WebSailor: Navigating Super-human Reasoning for Web Agent

Kuan Li, Zhongwang Zhang, Huifeng Yin, Liwen Zhang, Litu Ou, Jialong Wu, Wenbiao Yin, Baixuan Li, Zhengwei Tao, Xinyu Wang, Weizhou Shen, Junkai Zhang, Dingchu Zhang, Xixi Wu, Yong Jiang, Ming Yan, Pengjun Xie, Fei Huang, Jingren Zhou

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.CL、cs.AI

Metis-RISE: RL Incentivizes and SFT Enhances Multimodal Reasoning Model Learning

Haibo Qiu, Xiaohan Lan, Fanfan Liu, Xiaohu Sun, Delian Ruan, Peng Shi, Lin Ma

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.AI、cs.LG

Comments Project Page: https://github.com/MM-Thinking/Metis-RISE

Reward-Guided Speculative Decoding for Efficient LLM Reasoning

Baohao Liao, Yuhui Xu, Hanze Dong, Junnan Li, Christof Monz, Silvio Savarese, Doyen Sahoo, Caiming Xiong

机构 * Language Technology Lab, University of Amsterdam（阿姆斯特丹大学语言技术实验室） ； Salesforce AI Research（Salesforce人工智能研究） 纠错

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.CL、cs.AI

Comments 17 pages

e3: Learning to Explore Enables Extrapolation of Test-Time Compute for LLMs

Amrith Setlur, Matthew Y. R. Yang, Charlie Snell, Jeremy Greer, Ian Wu, Virginia Smith, Max Simchowitz, Aviral Kumar

专题命中 测试时计算 ：test-time compute(title)；reasoning(abstract)；分类 cs.CL、cs.LG

Structured Pruning for Diverse Best-of-N Reasoning Optimization

Hieu Trung Nguyen, Bao Nguyen, Viet Anh Nguyen

机构 * The Chinese University of Hong Kong（香港中文大学） 纠错

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.CL、cs.LG

Comments Accepted to ACL Findings 2025

ProRL: Prolonged Reinforcement Learning Expands Reasoning Boundaries in Large Language Models

Mingjie Liu, Shizhe Diao, Ximing Lu, Jian Hu, Xin Dong, Yejin Choi, Jan Kautz, Yi Dong

机构 * NVIDIA 纠错

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.CL、cs.AI

Comments 26 pages, 17 figures

Socratic-PRMBench: Benchmarking Process Reward Models with Systematic Reasoning Patterns

Xiang Li, Haiyang Yu, Xinghua Zhang, Ziyang Huang, Shizhu He, Kang Liu, Jun Zhao, Fei Huang, Yongbin Li

机构 * Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所） ； School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院） ； Tongyi Lab, Alibaba Group（阿里云实验室） 纠错

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.CL、cs.AI

Case-Based Reasoning Enhances the Predictive Power of LLMs in Drug-Drug Interaction

Guangyi Liu, Yongqi Zhang, Xunyuan Liu, Quanming Yao

机构 * Department of Electronic Engineering, Tsinghua University（清华大学电子工程系） ； The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州）） 纠错

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.AI、cs.LG

Temporal Sampling for Forgotten Reasoning in LLMs

Yuetai Li, Zhangchen Xu, Fengqing Jiang, Bhaskar Ramasubramanian, Luyao Niu, Bill Yuchen Lin, Xiang Yue, Radha Poovendran

机构 * University of Washington（华盛顿大学） ； Carnegie Mellon University（卡内基梅隆大学） ； Western Washington University（西华盛顿大学） 纠错

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.AI、cs.LG

The Unreasonable Effectiveness of Entropy Minimization in LLM Reasoning

Shivam Agarwal, Zimin Zhang, Lifan Yuan, Jiawei Han, Hao Peng

机构 * University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校） 纠错

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.AI、cs.LG

SpecReason: Fast and Accurate Inference-Time Compute via Speculative Reasoning

Rui Pan, Yinwei Dai, Zhihao Zhang, Gabriele Oliaro, Zhihao Jia, Ravi Netravali

机构 * Princeton University（普林斯顿大学） ； Carnegie Mellon University（卡内基梅隆大学） 纠错

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.AI、cs.LG

Aurelia: Test-time Reasoning Distillation in Audio-Visual LLMs

Sanjoy Chowdhury, Hanan Gani, Nishit Anand, Sayan Nag, Ruohan Gao, Mohamed Elhoseiny, Salman Khan, Dinesh Manocha

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.AI、cs.LG

ConSol: Sequential Probability Ratio Testing to Find Consistent LLM Reasoning Paths Efficiently

Jaeyeon Lee, Guantong Qi, Matthew Brady Neeley, Zhandong Liu, Hyun-Hwan Jeong

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.AI、cs.LG

Q-NL Verifier: Leveraging Synthetic Data for Robust Knowledge Graph Question Answering

Tim Schwabe, Louisa Siebel, Patrik Valach, Maribel Acosta

专题命中 测试时计算 ：verifier(title,abstract)；分类 cs.CL、cs.LG

Reasoning Aware Self-Consistency: Leveraging Reasoning Paths for Efficient LLM Sampling

Guangya Wan, Yuqi Wu, Jie Chen, Sheng Li

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.CL、cs.AI

Comments Accepted to NAACL 2025

StructRAG: Boosting Knowledge Intensive Reasoning of LLMs via Inference-time Hybrid Information Structurization

Zhuoqun Li, Xuanang Chen, Haiyang Yu, Hongyu Lin, Yaojie Lu, Qiaoyu Tang, Fei Huang, Xianpei Han, Le Sun, Yongbin Li

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.CL、cs.AI

AutoPSV: Automated Process-Supervised Verifier

Jianqiao Lu, Zhiyang Dou, Hongru Wang, Zeyu Cao, Jianbo Dai, Yingjia Wan, Zhijiang Guo

专题命中 测试时计算 ：verifier(title)；reasoning(abstract)；分类 cs.CL、cs.LG

Comments Accepted by NeurIPS 2024 Poster, 21 pages, 1 figure, 15 tables

TAP4LLM: Table Provider on Sampling, Augmenting, and Packing Semi-structured Data for Large Language Model Reasoning

Yuan Sui, Jiaru Zou, Mengyu Zhou, Xinyi He, Lun Du, Shi Han, Dongmei Zhang

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.CL、cs.AI

Comments This paper has been accepted by EMNLP 2024

Nuance Matters: Probing Epistemic Consistency in Causal Reasoning

Shaobo Cui, Junyou Li, Luca Mouchel, Yiyang Feng, Boi Faltings

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.CL、cs.AI

Comments 20 pages

MIDGARD: Self-Consistency Using Minimum Description Length for Structured Commonsense Reasoning

Inderjeet Nair, Lu Wang

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.CL、cs.AI

Comments Accepted at ACL 2024(main)

SMORE: Knowledge Graph Completion and Multi-hop Reasoning in Massive Knowledge Graphs

Hongyu Ren, Hanjun Dai, Bo Dai, Xinyun Chen, Denny Zhou, Jure Leskovec, Dale Schuurmans

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.AI、cs.LG

Turning Tables: Generating Examples from Semi-structured Tables for Endowing Language Models with Reasoning Skills

Ori Yoran, Alon Talmor, Jonathan Berant

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.CL、cs.LG

Deep Verifier Networks: Verification of Deep Discriminative Models with Deep Generative Models

Tong Che, Xiaofeng Liu, Site Li, Yubin Ge, Ruixiang Zhang, Caiming Xiong, Yoshua Bengio

专题命中 测试时计算 ：verifier(title,abstract)；分类 cs.AI、cs.LG

Comments Accepted to AAAI 2021

PhysNote: Self-Knowledge Notes for Evolvable Physical Reasoning in Vision-Language Model

PhysNote: 为视觉-语言模型中的可进化物理推理提供自我知识笔记

Sinin Zhang, Yunfei Xie, Yuxuan Cheng, Haoyu Zhang, Tong Zhang

机构 * The Chinese University of Hong Kong, Shenzhen（香港中文大学（深圳）） ； Rice University（里奇大学） ； City University of Hong Kong（香港城市大学） ； Fudan University（复旦大学） 纠错

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.AI

AI总结 本文提出PhysNote框架，通过自我生成的知识笔记提升视觉-语言模型在动态物理推理中的表现，实验显示其在PhysBench上达到56.68%的准确率，优于多智能体基线。

Comments 11 pages. Accepted by ICLR 2026 Workshop ES-Reasoning