大模型推理能力

Towards computing low-makespan solutions for multi-arm multi-task planning problems

Valentin N. Hartmann, Marc Toussaint

专题命中 规划推理 ：planning(title,abstract)

Comments Workshop for Planning and Robotics (PlanRob), International Conference on Automated Planning and Scheduling (ICAPS), 2023

Quantum Information Science and Technology for Nuclear Physics. Input into U.S. Long-Range Planning, 2023

Douglas Beck, Joseph Carlson, Zohreh Davoudi, Joseph Formaggio, Sofia Quaglioni, Martin Savage, Joao Barata, Tanmoy Bhattacharya, Michael Bishof, Ian Cloet, Andrea Delgado, Michael DeMarco, Caleb Fink, Adrien Florio, Marianne Francois, Dorota Grabowska, Shannon Hoogerheide, Mengyao Huang, Kazuki Ikeda, Marc Illa, Kyungseon Joo, Dmitri Kharzeev, Karol Kowalski, Wai Kin Lai, Kyle Leach, Ben Loer, Ian Low, Joshua Martin, David Moore, Thomas Mehen, Niklas Mueller, James Mulligan, Pieter Mumm, Francesco Pederiva, Rob Pisarski, Mateusz Ploskon, Sanjay Reddy, Gautam Rupak, Hersh Singh, Maninder Singh, Ionel Stetcu, Jesse Stryker, Paul Szypryt, Semeon Valgushev, Brent VanDevender, Samuel Watkins, Christopher Wilson, Xiaojun Yao, Andrei Afanasev, Akif Baha Balantekin, Alessandro Baroni, Raymond Bunker, Bipasha Chakraborty, Ivan Chernyshev, Vincenzo Cirigliano, Benjamin Clark, Shashi Kumar Dhiman, Weijie Du, Dipangkar Dutta, Robert Edwards, Abraham Flores, Alfredo Galindo-Uribarri, Ronald Fernando Garcia Ruiz, Vesselin Gueorguiev, Fanqing Guo, Erin Hansen, Hector Hernandez, Koichi Hattori, Philipp Hauke, Morten Hjorth-Jensen, Keith Jankowski, Calvin Johnson, Denis Lacroix, Dean Lee, Huey-Wen Lin, Xiaohui Liu, Felipe J. Llanes-Estrada, John Looney, Misha Lukin, Alexis Mercenne, Jeff Miller, Emil Mottola, Berndt Mueller, Benjamin Nachman, John Negele, John Orrell, Amol Patwardhan, Daniel Phillips, Stephen Poole, Irene Qualters, Mike Rumore, Thomas Schaefer, Jeremy Scott, Rajeev Singh, James Vary, Juan-Jose Galvez-Viruet, Kyle Wendt, Hongxi Xing, Liang Yang, Glenn Young, Fanyi Zhao

专题命中 规划推理 ：planning(title,abstract)

Comments A white paper for the 2023 nuclear physics long-range planning activity, emerging from the workshop "Quantum Information Science for U.S. Nuclear Physics Long Range Planning'', held in Santa Fe, New Mexico on January 31 - February 1, 2023. 26 pages with 7 figures

PiP: Planning-informed Trajectory Prediction for Autonomous Driving

Haoran Song, Wenchao Ding, Yuxuan Chen, Shaojie Shen, Michael Yu Wang, Qifeng Chen

专题命中 规划推理 ：planning(title,abstract)

Comments European Conference on Computer Vision (ECCV) 2020; Project page at http://haoran-song.github.io/planning-informed-prediction

Planning With Discrete Harmonic Potential Fields

Ahmad A. Masoud

专题命中 规划推理 ：planning(title,abstract)

Journal ref "Mobile Robots Motion Planning: New Challanges", I-TECH, Vienna, Austeria, 2008, Pp. 335-360

Planning in the Wild: Modeling Tools for PDDL

Volker Strobel, Alexandra Kirsch

专题命中 规划推理 ：planning(title,abstract)

Journal ref Strobel, Volker, and Alexandra Kirsch. "Planning in the Wild: Modeling Tools for PDDL." KI 2014: Advances in Artificial Intelligence. Springer International Publishing, 2014. 273-284

A frequentist framework of inductive reasoning

David R. Bickel

专题命中 规划推理 ：reasoning(title,abstract)

Comments major revision

Journal ref D. R. Bickel, "A frequentist framework of inductive reasoning," Sankhya A 74, 141-169 (2012)

Transformers Struggle to Use Their Emergent World Models: Revisiting the Tower of Hanoi, and the Illusion of Thinking

Transformer难以利用其涌现的世界模型：重新审视汉诺塔与思维的错觉

Devin Pereira, Willem Zuidema

机构 * University of Amsterdam（阿姆斯特丹大学） 纠错

专题命中 规划推理 ：reasoning(abstract)；chain-of-thought(abstract)；planning(abstract)；分类 cs.AI、cs.LG

AI总结 本文研究发现，Qwen3.6-27B等大型推理模型虽能编码汉诺塔的谢尔宾斯基世界模型，但因表示衰减导致圆环数超3时失败，注入提示词时间表示可部分恢复性能。

Long-Horizon Plan Execution in Large Tool Spaces through Entropy-Guided Branching

在大规模工具空间中通过熵引导分支实现长周期计划执行

Rongzhe Wei, Ge Shi, Min Cheng, Na Zhang, Pan Li, Sarthak Ghosh, Vaibhav Gorde, Leman Akoglu

机构 * Georgia Institute of Technology（佐治亚理工学院） ； Amazon（亚马逊） 纠错

专题命中 规划推理 ：reasoning(abstract)；planning(abstract)；self-correction(abstract)；分类 cs.CL、cs.AI

AI总结 本文提出熵引导分支算法，解决大规模工具库中长周期任务执行的挑战，通过动态扩展高预测熵的决策分支，提升任务成功率和计算效率。

Comments This work was completed during an internship at Amazon

FOR-Prompting: From Objection to Revision via an Asymmetric Prompting Protocol

从反对到修订的非对称提示协议：FOR-Prompting

He Zhang, Anzhou Zhang, Jian Dai

机构 * The Pennsylvania State University（宾夕法尼亚州立大学） ； Genentech, Inc.（基因泰克公司） 纠错

专题命中 规划推理 ：reasoning(abstract)；CoT(abstract)；planning(abstract)；分类 cs.CL、cs.AI

AI总结 FOR-Prompting通过非对称提示协议实现从反对到修订的推理机制，提升低资源环境下模型的推理能力与任务表现。

Time-Scaling Is What Agents Need Now

现在智能体需要时间扩展

Zhi Liu, Guangzhi Wang

机构 * CareerInternational Research Team（CareerInternational研究团队） 纠错

专题命中 规划推理 ：reasoning(abstract)；chain-of-thought(abstract)；CoT(abstract)；分类 cs.CL、cs.AI

AI总结 本文提出时间扩展概念，旨在通过扩展时间路径提升智能体的深度推理和问题解决能力，无需增加静态参数。

AppCopilot: Toward General, Accurate, Long-Horizon, and Efficient Mobile Agent

Jingru Fan, Yufan Dang, Jingyao Wu, Huatao Li, Runde Yang, Xiyuan Yang, Yuheng Wang, Chen Qian

专题命中 规划推理 ：reasoning(abstract)；chain-of-thought(abstract)；planning(abstract)；分类 cs.CL、cs.AI

Comments Project at https://github.com/OpenBMB/AppCopilot

Small Language Models for Agentic Systems: A Survey of Architectures, Capabilities, and Deployment Trade offs

Raghav Sharma, Manan Mehta

专题命中 规划推理 ：reasoning(abstract)；planning(abstract)；verifier(abstract)；分类 cs.AI、cs.LG

Comments 9 Pages

LLM-First Search: Self-Guided Exploration of the Solution Space

Nathan Herr, Tim Rocktäschel, Roberta Raileanu

专题命中 规划推理 ：reasoning(abstract)；planning(abstract)；test-time compute(abstract)；分类 cs.CL、cs.AI

Comments 9 main pages, 2 figures, 2 tables, 36 appendix pages

Control-R: Towards controllable test-time scaling

Di Zhang, Weida Wang, Junxian Li, Xunzhi Wang, Jiatong Li, Jianbo Wu, Jingdi Lei, Haonan He, Peng Ye, Shufei Zhang, Wanli Ouyang, Yuqiang Li, Dongzhan Zhou

机构 * Fudan University（复旦大学） ； Shanghai Artificial Intelligence Laboratory（上海人工智能实验室） ； Tongji University（同济大学） ； Shanghai Jiaotong University（上海交通大学） ； Nankai University（南开大学） ； Hong Kong Polytechnic University（香港理工大学） ； University of California, Merced（加州大学默塞德分校） ； University of Science and Technology of China（中国科学技术大学） 纠错

专题命中 规划推理 ：reasoning(abstract)；chain-of-thought(abstract)；CoT(abstract)；分类 cs.CL、cs.AI

Effects of Theory of Mind and Prosocial Beliefs on Steering Human-Aligned Behaviors of LLMs in Ultimatum Games

Neemesh Yadav, Palakorn Achananuparp, Jing Jiang, Ee-Peng Lim

机构 * Singapore Management University（新加坡管理大学） ； Australian National University（澳大利亚国立大学） 纠错

专题命中 规划推理 ：reasoning(abstract)；chain-of-thought(abstract)；CoT(abstract)；分类 cs.CL、cs.AI

Comments 17 pages, 1 figure, 6 tables

Performance of Large Language Models in Supporting Medical Diagnosis and Treatment

Diogo Sousa, Guilherme Barbosa, Catarina Rocha, Dulce Oliveira

专题命中 规划推理 ：reasoning(abstract)；chain-of-thought(abstract)；planning(abstract)；分类 cs.CL、cs.AI

Comments 21 pages, 6 figures, 4 tables. Acknowledgements: The authors acknowledge the support of the AITriage4SU Project (2024.07400.IACDC/2024), funded by the FCT (Foundation for Science and Technology), Portugal

Explainable LLM-driven Multi-dimensional Distillation for E-Commerce Relevance Learning

Gang Zhao, Ximing Zhang, Chenji Lu, Hui Zhao, Tianshu Wu, Pengjie Wang, Jian Xu, Bo Zheng

专题命中 规划推理 ：reasoning(abstract)；chain-of-thought(abstract)；CoT(abstract)；分类 cs.CL、cs.AI

Comments Accepted by WWW 2025 oral

Toward Self-Improvement of LLMs via Imagination, Searching, and Criticizing

Ye Tian, Baolin Peng, Linfeng Song, Lifeng Jin, Dian Yu, Haitao Mi, Dong Yu

专题命中 规划推理 ：reasoning(abstract)；planning(abstract)；self-correction(abstract)；分类 cs.CL、cs.LG

Comments NeurIPS 2024

MageBench: Bridging Large Multimodal Models to Agents

Miaosen Zhang, Qi Dai, Yifan Yang, Jianmin Bao, Dongdong Chen, Kai Qiu, Chong Luo, Xin Geng, Baining Guo

专题命中 规划推理 ：reasoning(abstract)；chain-of-thought(abstract)；planning(abstract)；分类 cs.AI、cs.LG

Comments 37 pages, 32 figures, github link: https://github.com/microsoft/MageBench

A Language Agent for Autonomous Driving

Jiageng Mao, Junjie Ye, Yuxi Qian, Marco Pavone, Yue Wang

专题命中 规划推理 ：reasoning(abstract)；chain-of-thought(abstract)；planning(abstract)；分类 cs.CL、cs.AI

Comments COLM 2024. Project Page: https://usc-gvl.github.io/Agent-Driver/

HANS, are you clever? Clever Hans Effect Analysis of Neural Systems

Leonardo Ranaldi, Fabio Massimo Zanzotto

专题命中 规划推理 ：reasoning(abstract)；chain-of-thought(abstract)；CoT(abstract)；分类 cs.CL、cs.AI

Comments This paper contains erroneous evaluations and we would like to withdraw it

Empowering Autonomous Driving with Large Language Models: A Safety Perspective

Yixuan Wang, Ruochen Jiao, Sinong Simon Zhan, Chengtian Lang, Chao Huang, Zhaoran Wang, Zhuoran Yang, Qi Zhu

专题命中 规划推理 ：reasoning(abstract)；planning(abstract)；verifier(abstract)；分类 cs.AI、cs.LG

Comments Accepted to LLMAgent workshop @ICLR2024

REX: Rapid Exploration and eXploitation for AI Agents

Rithesh Murthy, Shelby Heinecke, Juan Carlos Niebles, Zhiwei Liu, Le Xue, Weiran Yao, Yihao Feng, Zeyuan Chen, Akash Gokul, Devansh Arpit, Ran Xu, Phil Mui, Huan Wang, Caiming Xiong, Silvio Savarese

专题命中 规划推理 ：reasoning(abstract)；CoT(abstract)；planning(abstract)；分类 cs.AI、cs.LG

OlaGPT: Empowering LLMs With Human-like Problem-Solving Abilities

Yuanzhen Xie, Tao Xie, Mingxiong Lin, WenTao Wei, Chenglin Li, Beibei Kong, Lei Chen, Chengxiang Zhuo, Bo Hu, Zang Li

专题命中 规划推理 ：reasoning(abstract)；chain-of-thought(abstract)；CoT(abstract)；分类 cs.CL、cs.AI

Think-to-Personalize: Unifying Reasoning and Retrieval for User-Centric Personalized Dense Retrieval

思考以个性化：统一推理与检索的以用户为中心的个性化密集检索

Angqing Jiang, Gaoming Zhang, Jianchun Song, Kena Qi, Dayao Chen, Wei Lin, Defu Lian

专题命中 规划推理 ：reasoning(title,abstract)

AI总结 本文提出TTP框架，将用户意图推理与密集检索统一，经两阶段训练后在电商场景实验及在线测试中均取得优于基线的效果，提升了订单量。

Comments Accepted at CIKM 2026. 11 pages, 8 figures, and 9 tables

Capability-Based Planning for AI Crisis Preparedness

面向AI危机准备的基于能力的规划

Isaak Mengesha, Charlie Collins, Juan Felipe Cerón Uribe, Salvatore d'Ambrosio, Vickie Ellis

专题命中 规划推理 ：planning(title,abstract)

AI总结 该研究针对AI难以预测的问题，提出三部分组成的基于能力的规划方法框架，经四类AI赋能威胁试点验证，为AI危机准备提供实用工具。

DELIVER: A System for LLM-Guided Coordinated Multi-Robot Pickup and Delivery using Voronoi-Based Relay Planning

Alkesh K. Srivastava, Jared Michael Levin, Alexander Derrico, Philip Dames

机构 * Department of Mechanical Engineering, Temple University（机械工程系， Temple 大学） 纠错

专题命中 规划推理 ：planning(title,abstract)

Comments Submission under review at the 2026 IEEE/SICE International Symposium on System Integration (SII 2026)

Planning Against Learning in Rank-1 Games

在秩-1博弈中针对学习的规划

William Overman

专题命中 规划推理 ：planning(title,abstract)

AI总结 该研究针对满足$\text{rank}(A+B)=1$的双矩阵博弈，证明其纳什均衡可多项式时间计算，但针对学习动态的最优规划是NP困难的，区分了高效均衡计算与策略规划。

MMDynOpt-Agent: Dynamic Optimization for Multimodal Large Language Model Reasoning via Reinforcement Learning

MMDynOpt-Agent：基于强化学习的多模态大语言模型推理动态优化方法

Wenjin Liu, Haoran Luo, Fayuan Ke, Zhenghong Lin, Yue Lu, Zhe Cui, Anh Tuan Luu, Carl Yang

专题命中 规划推理 ：reasoning(title,abstract)

AI总结 该研究针对多模态大语言模型推理效率不足的问题，提出MMDynOpt-Agent，通过强化学习将多模态推理动态优化建模为马尔可夫决策过程，结合多轮优化提示与多维度奖励机制，在15个公开数据集上性能优于基线方法。

HxAgent: Iterative Agent Planning for End-to-End Web Application Testing

HxAgent：用于端到端Web应用测试的迭代智能体规划方法

Tu Nguyen, Duy Cao, Viet Nguyen, Phu Nguyen, Vy Le, Nguyen TK Nguyen, Tien N. Nguyen, Vu Nguyen

专题命中 规划推理 ：planning(title,abstract)

AI总结 HxAgent是一种基于LLM的迭代规划智能体，通过主动修正策略结合多类信息优化Web测试，在MiniWoB++、350项Web任务及OnlineMind2Web数据集上均优于WALT模型，表现优异。

Comments Under review for a conference