大模型推理能力

ThinkDial: An Open Recipe for Controlling Reasoning Effort in Large Language Models

Qianyu He, Siyu Yuan, Xuefeng Li, Mingxuan Wang, Jiangjie Chen

机构 * Fudan University（复旦大学） ； Shanghai Jiao Tong University（上海交通大学） ； SIA-Lab of Tsinghua AIR（清华大学AIR实验室） 纠错

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

Long-context Language Models Fail in Basic Retrieval Tasks Without Sufficient Reasoning Steps

Yijiong Yu, Yongfeng Huang, Zhixiao Qi, Wei Wang, Weifeng Liu, Ran Chen, Ji Pei

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

Comments Our code is publicly available at https://github.com/yuyijiong/hard_retrieval_for_llm and the datasets is at https://huggingface.co/datasets/yuyijiong/difficult_retrieval

Architecting Clinical Collaboration: Multi-Agent Reasoning Systems for Multimodal Medical VQA

Karishma Thakrar, Shreyas Basavatia, Akshay Daftardar

机构 * Georgia Institute of Technology Atlanta, GA, USA（佐治亚理工学院） 纠错

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

Accelerate Parallelizable Reasoning via Parallel Decoding within One Sequence

Yijiong Yu

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

Comments Our code is available in https://github.com/yuyijiong/parallel-decoding-in-one-sequence

Dynamic Collaboration of Multi-Language Models based on Minimal Complete Semantic Units

Chao Hao, Zezheng Wang, Yanhua Huang, Ruiwen Xu, Wenzhe Niu, Xin Liu, Zitong Yu

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

Comments Accepted by EMNLP 2025 Main Conference

History Rhymes: Accelerating LLM Reinforcement Learning with RhymeRL

Jingkai He, Tianjian Li, Erhu Feng, Dong Du, Qian Liu, Tao Liu, Yubin Xia, Haibo Chen

机构 * Institute of Parallel and Distributed Systems, Shanghai Jiao Tong University（并行与分布式系统研究所，上海交通大学） ； ByteDance（字节跳动） 纠错

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

The AI in the Mirror: LLM Self-Recognition in an Iterated Public Goods Game

Olivia Long, Carter Teplica

机构 * Columbia University（哥伦比亚大学） ； Mila, Polytechnique Montréal（蒙特利尔大学Mila） 纠错

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

Integral Transformer: Denoising Attention, Not Too Much Not Too Little

Ivan Kobyzev, Abbas Ghaddar, Dingtao Hu, Boxing Chen

机构 * Huawei Noah’s Ark Lab, Montreal Research Center, Canada（华为诺亚实验室，蒙特利尔研究中心，加拿大） 纠错

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

Comments EMNLP 2025 Main

M$^2$IV: Towards Efficient and Fine-grained Multimodal In-Context Learning via Representation Engineering

Yanshu Li, Yi Cao, Hongyang He, Qisen Cheng, Xiang Fu, Xi Xiao, Tianyang Wang, Ruixiang Tang

机构 * Brown University（布朗大学） ； University of Warwick（沃里克大学） ； Samsung US（三星美国分公司） ； Boston University（波士顿大学） ； University of Alabama at Birmingham（阿拉巴马大学伯明翰分校） ； Rutgers University（罗格斯大学） 纠错

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

Comments COLM 2025, 30 pages, 10 figures, 16 tables

General Intelligence Requires Reward-based Pretraining

Seungwook Han, Jyothish Pari, Samuel J. Gershman, Pulkit Agrawal

机构 * Improbable AI Lab, MIT（Improbable AI Lab，麻省理工学院） ； Department of Psychology（心理学系） ； Center for Brain Science, Harvard University（哈佛大学脑科学中心） 纠错

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

Comments https://improbableai.notion.site/General-Intelligence-Requires-Reward-Based-Pretraining-2023b66e4cf580d3ab44c7860b75d25f?pvs=73

Toward Edge General Intelligence with Agentic AI and Agentification: Concepts, Technologies, and Future Directions

Ruichen Zhang, Guangyuan Liu, Yinqiu Liu, Changyuan Zhao, Jiacheng Wang, Yunting Xu, Dusit Niyato, Jiawen Kang, Yonghui Li, Shiwen Mao, Sumei Sun, Xuemin Shen, Dong In Kim

专题命中 其他推理 ：reasoning(abstract)

Skeptik: A Hybrid Framework for Combating Potential Misinformation in Journalism

Arlen Fan, Fan Lei, Steven R. Corman, Ross Maciejewski

专题命中 其他推理 ：reasoning(abstract)

Comments Arlen Fan and Fan Lei contributed equally to this research. Accepted by ACM Transactions on Interactive Intelligent Systems (TiiS)