arXivDaily arXiv每日学术速递 周一至周五更新

AI 大模型

大模型推理能力

大模型数学、逻辑、规划、多步推理和测试时计算能力。

共收录 5757 信号源:cs.CL, cs.AI, cs.LG

1. 其他推理 5757 篇

2608.01980 2026-08-04 cs.CV cs.AI 新提交 90%

AdaThinkV: Adaptive Thinking for Token-Efficient Video Reasoning

AdaThinkV:面向令牌高效视频推理的自适应思维

Jingqi Tian, Haoji Zhang, Lin Chen, Hongbo Jin, Haonan Xu, Tianrui Zhu, Xingming Shui, Shilin Ma, Wenjing Yang, Yansong Tang

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

AI总结 本研究提出AdaThinkV自适应视频推理框架,通过ThinkGain与VRPO解决CoT推理的令牌浪费问题,在视频推理任务中以更少令牌实现优于最强自适应基线的准确率。

详情

展开后加载摘要…

URL PDF HTML 收藏
2607.27304 2026-07-31 cs.LG cs.CV 新提交 90%

Position, Not Provenance: Separating Reasoning Mediation from Sycophancy in Medical Vision-Language Models

位置,而非来源:在医学视觉-语言模型中分离推理中介与逢迎行为

Supratik Bhowal, Subhrajyoti Basu, Aritra Gir Mahanta, Anik Pal Chowdhury

机构 * IEM Kolkata(印度工程管理学院 Kolkata校区) School of UEMK Kolkata(UEMK Kolkata学院) Heritage Institute of Technology Kolkata(加尔各答遗产技术学院) Indian Institute of Information Technology, Kalyani(卡利亚尼印度信息技术学院)

专题命中 其他推理 :CoT(summary_cn,abstract);reasoning(title,abstract);chain-of-thought(abstract);分类 cs.LG

AI总结 本研究提出CoT-Mediate框架,结合双臂协议与来源控制干预,在VQA-RAD数据集上评估LLaVA-Med和MedGemma,发现上下文位置而非声明来源是医学VLMs使用生成推理的主要决定因素。

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.27025 2026-06-26 cs.CL 新提交 90%

Improving General Role-Playing Agents via Psychology-Grounded Reasoning and Role-Aware Policy Optimization

通过心理学推理和角色感知策略优化改进通用角色扮演智能体

Zhenhua Xu, Dongsheng Chen, Jian Li, Yitong Lin, Zhebo Wang, Jiafu Wu, Yizhang Jin, Chengjie Wang, Meng Han, Yabiao Wang

机构 * Zhejiang University(浙江大学) Tencent Youtu Lab(腾讯优图实验室)

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

AI总结 提出Psy-CoT框架,将角色推理分解为三步,并结合RAPO算法通过互信息加权梯度,提升角色扮演的忠实度和泛化能力。

详情

展开后加载摘要…

URL PDF HTML 收藏
2605.28600 2026-05-28 cs.LG 90%

Transformers Provably Learn to Internalize Chain-of-Thought

Transformer可证明地内化思维链

Yixiao Huang, Hanlin Zhu, Zixuan Wang, Jiantao Jiao, Stuart Russell, Somayeh Sojoudi, Song Mei

机构 * UC Berkeley(伯克利大学) Princeton University(普林斯顿大学)

专题命中 其他推理 :CoT(summary_cn,abstract);chain-of-thought(title,abstract);reasoning(abstract);分类 cs.LG

AI总结 提出Log-ICoT课程训练,使L层Transformer用多项式样本学习k-奇偶性,匹配显式CoT的样本效率且消除推理开销。

详情

展开后加载摘要…

URL PDF HTML 收藏
2603.00578 2026-03-03 cs.AI cs.CL 90%

Draft-Thinking: Learning Efficient Reasoning in Long Chain-of-Thought LLMs

草稿式推理:在长链式推理LLM中学习高效推理

Jie Cao, Tianwei Lin, Zhenxuan Fan, Bo Yuan, Ziyuan Zhao, Rolan Yan, Wenqiao Zhang, Siliang Tang

机构 * Zhejiang University(浙江大学) Tencent(腾讯)

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

AI总结 Draft-Thinking通过草稿式推理结构和渐进式课程学习,在长链式推理LLM中实现高效推理,显著降低推理预算并保持性能

详情

展开后加载摘要…

URL PDF HTML 收藏
2505.23575 2025-11-26 cs.AI cs.LG 90%

CoT Red-Handed: Stress Testing Chain-of-Thought Monitoring

CoT红手:链式推理监控的压力测试

Benjamin Arnav, Pablo Bernabeu-Pérez, Nathan Helm-Burger, Tim Kostolansky, Hannes Whittingham, Mary Phuong

机构 * LASR Labs(语言与语音研究实验室)

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

AI总结 本文提出了一种混合协议,通过结合模型推理和行动评分,提升链式推理监控在检测有害行为方面的性能。

Comments To be published in the 39th Conference on Neural Information Processing Systems (NeurIPS 2025)

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.07378 2025-11-11 cs.LG cs.AI math.OC stat.ML 90%

Transformers Provably Learn Chain-of-Thought Reasoning with Length Generalization

Yu Huang, Zixin Wen, Aarti Singh, Yuejie Chi, Yuxin Chen

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

Comments This is the full version of a paper published at NeurIPS 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
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

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.04045 2025-10-07 cs.CL cs.LG 90%

Exploring Chain-of-Thought Reasoning for Steerable Pluralistic Alignment

Yunfan Zhang, Kathleen McKeown, Smaranda Muresan

机构 * Columbia University(哥伦比亚大学) Barnard College(巴纳德学院)

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

Comments ACL EMNLP 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.20317 2025-09-26 cs.CL cs.AI 90%

SIM-CoT: Supervised Implicit Chain-of-Thought

Xilin Wei, Xiaoran Liu, Yuhang Zang, Xiaoyi Dong, Yuhang Cao, Jiaqi Wang, Xipeng Qiu, Dahua Lin

机构 * Fudan University(复旦大学) Shanghai AI Laboratory(上海人工智能实验室) The Chinese University of Hong Kong(香港中文大学) Shanghai Innovation Institute(上海创新研究院)

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2507.23751 2025-09-04 cs.AI cs.CL 90%

CoT-Self-Instruct: Building high-quality synthetic prompts for reasoning and non-reasoning tasks

Ping Yu, Jack Lanchantin, Tianlu Wang, Weizhe Yuan, Olga Golovneva, Ilia Kulikov, Sainbayar Sukhbaatar, Jason Weston, Jing Xu

机构 * FAIR at Meta(Meta的FAIR)

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2508.00574 2025-08-04 cs.CL cs.AI 90%

SynAdapt: Learning Adaptive Reasoning in Large Language Models via Synthetic Continuous Chain-of-Thought

Jianwei Wang, Ziming Wu, Fuming Lai, Shaobing Lian, Ziqian Zeng

机构 * Tencent Inc.(腾讯公司)

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2506.12301 2025-06-17 cs.LG cs.AI 90%

Unveiling Confirmation Bias in Chain-of-Thought Reasoning

Yue Wan, Xiaowei Jia, Xiang Lorraine Li

机构 * University of Pittsburgh(匹兹堡大学)

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

Journal ref ACL 2025 Findings

详情

展开后加载摘要…

URL PDF HTML 收藏
2505.21784 2025-05-29 cs.AI cs.CL 90%

Towards Safety Reasoning in LLMs: AI-agentic Deliberation for Policy-embedded CoT Data Creation

Tharindu Kumarage, Ninareh Mehrabi, Anil Ramakrishna, Xinyan Zhao, Richard Zemel, Kai-Wei Chang, Aram Galstyan, Rahul Gupta, Charith Peris

机构 * Amazon Nova Responsible AI(亚马逊Nova负责任的人工智能) Arizona State University(亚利桑那州立大学)

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

Comments Accepted to ACL 2025 (Findings)

详情

展开后加载摘要…

URL PDF HTML 收藏
2505.10185 2025-05-16 cs.CL cs.AI 90%

The CoT Encyclopedia: Analyzing, Predicting, and Controlling how a Reasoning Model will Think

Seongyun Lee, Seungone Kim, Minju Seo, Yongrae Jo, Dongyoung Go, Hyeonbin Hwang, Jinho Park, Xiang Yue, Sean Welleck, Graham Neubig, Moontae Lee, Minjoon Seo

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

Comments Work in progress

详情

展开后加载摘要…

URL PDF HTML 收藏
2504.16913 2025-04-24 cs.CL cs.AI 90%

Tracing Thought: Using Chain-of-Thought Reasoning to Identify the LLM Behind AI-Generated Text

Shifali Agrahari, Sanasam Ranbir Singh

机构 * Department of Computer Science and Engineering(计算机科学与工程系) Indian Institute of Technology Guwahati(印度理工学院瓜哇蒂)

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 De-Factify 4: 4th Workshop on Multimodal Fact Checking and Hate Speech Detection, co-located with AAAI 2025. Pennsylvania

详情

展开后加载摘要…

URL PDF HTML 收藏
2502.10428 2025-04-08 cs.AI cs.LG 90%

Dynamic Chain-of-Thought: Towards Adaptive Deep Reasoning

Libo Wang

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

Comments The GitHub repository link is: https://github.com/brucewang123456789/GeniusTrail/tree/main/Dynamic%20CoT

详情

展开后加载摘要…

URL PDF HTML 收藏
2405.20902 2024-06-03 cs.CL cs.AI cs.CR 90%

Preemptive Answer "Attacks" on Chain-of-Thought Reasoning

Rongwu Xu, Zehan Qi, Wei Xu

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 Accepted to ACL'24 (Findings). Camera-ready version

详情

展开后加载摘要…

URL PDF HTML 收藏
2402.18312 2024-05-07 cs.CL cs.LG 90%

How to think step-by-step: A mechanistic understanding of chain-of-thought reasoning

Subhabrata Dutta, Joykirat Singh, Soumen Chakrabarti, Tanmoy Chakraborty

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2403.19167 2024-03-29 cs.CL cs.AI 90%

Mitigating Misleading Chain-of-Thought Reasoning with Selective Filtering

Yexin Wu, Zhuosheng Zhang, Hai Zhao

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2401.07037 2024-01-17 cs.CL cs.AI 90%

xCoT: Cross-lingual Instruction Tuning for Cross-lingual Chain-of-Thought Reasoning

Linzheng Chai, Jian Yang, Tao Sun, Hongcheng Guo, Jiaheng Liu, Bing Wang, Xiannian Liang, Jiaqi Bai, Tongliang Li, Qiyao Peng, Zhoujun Li

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 11 pages

详情

展开后加载摘要…

URL PDF HTML 收藏
2308.13259 2023-10-31 cs.CL cs.AI 90%

Knowledge-Driven CoT: Exploring Faithful Reasoning in LLMs for Knowledge-intensive Question Answering

Keheng Wang, Feiyu Duan, Sirui Wang, Peiguang Li, Yunsen Xian, Chuantao Yin, Wenge Rong, Zhang Xiong

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2310.16535 2023-10-26 cs.CL cs.AI 90%

R$^3$ Prompting: Review, Rephrase and Resolve for Chain-of-Thought Reasoning in Large Language Models under Noisy Context

Qingyuan Tian, Hanlun Zhu, Lei Wang, Yang Li, Yunshi Lan

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2310.14799 2023-10-24 cs.CL cs.AI 90%

Cross-lingual Prompting: Improving Zero-shot Chain-of-Thought Reasoning across Languages

Libo Qin, Qiguang Chen, Fuxuan Wei, Shijue Huang, Wanxiang Che

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 Accepted at EMNLP2023 Main Conference

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.22158 2026-07-01 cs.CV 新提交 89%

Improving Reasoning in Vision-Language Models via Perception Verified Self-Training

通过感知验证自训练提升视觉-语言模型的推理能力

Sourabh Sharma, Sonam Gupta, Sadbhawna

机构 * Malaviya National Institute of Technology Jaipur(马拉维亚国家理工学院斋浦尔分校) IBM Research(IBM研究院)

专题命中 其他推理 :CoT(summary_cn,abstract);reasoning(title,abstract);chain-of-thought(abstract)

AI总结 提出感知验证自训练框架,通过解耦感知与推理的CoT模板和无监督评估方法PerceptEval,结合两阶段课程学习,提升视觉-语言模型在视觉推理中的准确性并减少幻觉。

Comments Accepted at The European Conference on Computer Vision 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.14961 2026-06-16 cs.CL 新提交 89%

CoRA: Confidence-Rationale Alignment for Reliable Chain-of-Thought Reasoning

CoRA: 面向可靠思维链推理的置信度-理由对齐

Juming Xiong, Weixin Liu, Kevin Guo, Congning Ni, Junchao Zhu, Chongyu Qu, Chao Yan, Katherine Brown, Avinash Baidya, Xiang Gao, Bradley Malin, Zhijun Yin

机构 * Vanderbilt University(范德比尔特大学) Vanderbilt University Medical Center(范德比尔特大学医学中心) Intuit AI Research(Intuit AI研究)

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

AI总结 提出GRPO强化学习框架,联合奖励答案正确性、置信度与理由支持度,减少置信度与理由对齐误差,提升推理可靠性。

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.13879 2026-03-12 cs.MM cs.CL cs.CV 89%

Chain-of-Thought Compression Should Not Be Blind: V-Skip for Efficient Multimodal Reasoning via Dual-Path Anchoring

链式推理压缩不应盲目:通过双路径锚定实现高效的多模态推理的V-Skip

Dongxu Zhang, Yiding Sun, Cheng Tan, Wenbiao Yan, Ning Yang, Jihua Zhu, Haijun Zhang

机构 * School of Software Engineering, Xi’an Jiaotong University(西安交通大学软件工程学院) Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所) Shanghai Artificial Intelligence Laboratory(上海人工智能实验室) Harbin Institute of Technology, Shenzhen(深圳哈尔滨工业大学) University of Science and Technology Beijing(北京科技大学)

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

AI总结 V-Skip通过双路径锚定机制解决多模态推理中令牌剪枝的盲目性问题,实现高效的推理速度提升与精度保持。

详情

展开后加载摘要…

URL PDF HTML 收藏
2602.14903 2026-02-26 cs.AI 89%

The Potential of CoT for Reasoning: A Closer Look at Trace Dynamics

链式推理在推理中的潜力:对轨迹动态的深入考察

Gregor Bachmann, Yichen Jiang, Seyed Mohsen Moosavi Dezfooli, Moin Nabi

机构 * Apple(苹果公司)

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

AI总结 本文研究了链式推理(CoT)在推理中的潜力,通过分析轨迹动态,发现CoT中不同部分对最终答案的贡献,并探讨了CoT的可转移性。

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.08641 2026-02-06 cs.AI q-fin.TR 89%

Resisting Manipulative Bots in Meme Coin Copy Trading: A Multi-Agent Approach with Chain-of-Thought Reasoning

抵制操纵机器人在表情包加密货币复制交易中的应用:一种基于多智能体的链式推理方法

Yichen Luo, Yebo Feng, Jiahua Xu, Yang Liu

机构 * UCL, Centre for Blockchain Technologies(伦敦大学区块链技术中心) The University of Hong Kong, FinTech Academy(香港大学金融科技学院) Nanyang Technological University(南洋理工大学)

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

AI总结 本文提出一种基于多智能体和链式推理的复制交易系统,以抵御操纵机器人,通过多模态大语言模型提升预测准确度和经济表现,实现加密货币投资的稳健收益。

Journal ref Proceedings of the ACM Web Conference 2026 (WWW'26)

详情

展开后加载摘要…

URL PDF HTML 收藏
2506.16123 2026-02-03 cs.CL 89%

FinCoT: Grounding Chain-of-Thought in Expert Financial Reasoning

FinCoT:在专家金融推理中 grounding 思维链

Natapong Nitarach, Warit Sirichotedumrong, Panop Pitchayarthorn, Pittawat Taveekitworachai, Potsawee Manakul, Kunat Pipatanakul

机构 * SCB 10X, SCBX Group(SCB 10X集团)

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

AI总结 FinCoT通过整合专家金融推理蓝图,提升金融领域模型性能并减少推理成本,实现更可解释的推理过程。

Comments Accepted at FinNLP-2025, EMNLP (Oral Presentation)

详情

展开后加载摘要…

URL PDF HTML 收藏