机器人 / 具身智能 - arXivDaily 专题

2509.07593 2026-02-03 cs.RO cs.AI cs.CV cs.SY eess.IV eess.SY 67%

Vision-Proprioception Fusion with Mamba2 in End-to-End Reinforcement Learning for Motion Control

基于Mamba2的视觉-本体感知融合在端到端强化学习中的运动控制

Xiaowen Tao, Yinuo Wang, Jinzhao Zhou

机构 * School of Computer Science and Statistics, Trinity College Dublin（都柏林三一学院计算机科学与统计学系）； Faculty of Engineering and Information Technology, University of Technology Sydney（新南威尔士大学理工学院）

专题命中模仿学习与强化学习：robotic(abstract)；分类 cs.RO、cs.AI、cs.CV

AI总结本文提出基于SSD-Mamba2的视觉-本体感知融合框架，通过端到端强化学习提升运动控制的效率和安全性。

Comments 6 figures and 8 tables. This paper has been accepted by Advanced Engineering Informatics

Journal ref Advanced Engineering Informatics, vol. 71, 2026

2601.04287 2026-01-12 cs.LG cs.AI cs.MA cs.RO 67%

Online Action-Stacking Improves Reinforcement Learning Performance for Air Traffic Control

在线动作堆叠提升空交通管制的强化学习性能

Ben Carvell, George De Ath, Eseoghene Benjamin, Richard Everson

机构 * Project Bluebird, NATS（Project Bluebird，NATS）； Department of Computer Science, University of Exeter（计算机科学系，埃克塞特大学）； Project Bluebird, The Alan Turing Institute（Project Bluebird，艾伦·图灵研究所）； The Alan Turing Institute and the University of Exeter（艾伦·图灵研究所和埃克塞特大学）

专题命中模仿学习与强化学习：navigation(abstract)；分类 cs.RO、cs.AI、cs.LG

AI总结在线动作堆叠通过简化动作空间提升空交通管制的强化学习性能，有效减少指令数量并实现与复杂动作空间相当的控制效果。

2505.14975 2026-01-05 cs.LG cs.AI cs.RO 67%

Flattening Hierarchies with Policy Bootstrapping

通过策略自举 flattening 层次结构

John L. Zhou, Jonathan C. Kao

机构 * University of California, Los Angeles（加州大学洛杉矶分校）

专题命中模仿学习与强化学习：manipulation(abstract)；分类 cs.RO、cs.AI、cs.LG

AI总结本文提出一种通过自举子目标策略训练平坦目标条件策略的方法，解决长horizon任务中GCRL扩展难题，实现高维控制性能提升。

Comments NeurIPS 2025 (Spotlight, top 3.2%)

2502.01956 2025-12-22 cs.RO cs.AI cs.LG 67%

DHP: Discrete Hierarchical Planning for Hierarchical Reinforcement Learning Agents

DHP：用于分层强化学习代理的离散分层规划

Shashank Sharma, Janina Hoffmann, Vinay Namboodiri

机构 * Department of Computer Science University of Bath（计算机科学系巴斯大学）； Department of Psychology University of Bath（心理学系巴斯大学）

专题命中模仿学习与强化学习：navigation(abstract)；分类 cs.RO、cs.AI、cs.LG

AI总结 DHP通过离散可达性检查和递归分解任务，提升分层强化学习代理在长距离视觉规划中的效率和泛化能力。

2512.00453 2025-12-02 cs.RO cs.AI cs.LG 67%

Sample-Efficient Expert Query Control in Active Imitation Learning via Conformal Prediction

通过置信预测实现高效的专家查询控制在主动模仿学习中

Arad Firouzkouhi, Omid Mirzaeedodangeh, Lars Lindemann

机构 * Department of Computer Science, University of Southern California（计算机科学系，南加州大学）； Department of Information Technology and Electrical Engineering, ETH Zurich（信息科技与电气工程系，苏黎世联邦理工学院）

专题命中模仿学习与强化学习：robotics(abstract)；分类 cs.RO、cs.AI、cs.LG

AI总结 CRSAIL通过置信预测实现高效的专家查询控制，在主动模仿学习中减少专家查询次数，提升机器人任务的训练效率。

2511.16593 2025-11-21 cs.SE cs.AI cs.CV cs.RO 67%

Green Resilience of Cyber-Physical Systems: Doctoral Dissertation

物理信息系统的绿色韧性：博士论文

Diaeddin Rimawi

专题命中模仿学习与强化学习：robot learning(abstract)；分类 cs.RO、cs.AI、cs.CV

AI总结本研究提出GResilience框架，通过多目标优化、博弈论和强化学习优化OL-CAIS的绿色恢复，减少能源影响并维持性能稳定性。

2511.12751 2025-11-18 cs.LG cs.AI cs.RO 67%

Are LLMs The Way Forward? A Case Study on LLM-Guided Reinforcement Learning for Decentralized Autonomous Driving

Timur Anvar, Jeffrey Chen, Yuyan Wang, Rohan Chandra

专题命中模仿学习与强化学习：navigation(abstract)；分类 cs.RO、cs.AI、cs.LG

2503.09956 2025-11-17 cs.LG cs.AI cs.CV cs.ET 67%

DeepSeek-Inspired Exploration of RL-based LLMs and Synergy with Wireless Networks: A Survey

Yu Qiao, Phuong-Nam Tran, Ji Su Yoon, Loc X. Nguyen, Eui-Nam Huh, Dusit Niyato, Choong Seon Hong

机构 * Kyung Hee University（韩国庆熙大学）； Nanyang Technological University（南洋理工大学）

专题命中模仿学习与强化学习：embodied AI(abstract)；分类 cs.AI、cs.CV、cs.LG

Comments 45 pages, 12 figures

Journal ref ACM Computing Surveys, Nov. 2025

2503.19037 2025-11-13 cs.LG cs.AI cs.RO 67%

Evolutionary Policy Optimization

Jianren Wang, Yifan Su, Abhinav Gupta, Deepak Pathak

专题命中模仿学习与强化学习：manipulation(abstract)；分类 cs.RO、cs.AI、cs.LG

Comments Website at https://yifansu1301.github.io/EPO/

2509.15981 2025-11-03 cs.LG cs.AI cs.RO stat.ML 67%

Uncertainty-Based Smooth Policy Regularisation for Reinforcement Learning with Few Demonstrations

Yujie Zhu, Charles A. Hepburn, Matthew Thorpe, Giovanni Montana

机构 * Department of Statistics（统计系）； Warwick Manufacturing Group University of Warwick（沃里克大学沃里克制造集团）

专题命中模仿学习与强化学习：robotics(abstract)；分类 cs.RO、cs.AI、cs.LG

2406.16258 2025-10-27 cs.RO cs.AI cs.LG 67%

MEReQ: Max-Ent Residual-Q Inverse RL for Sample-Efficient Alignment from Intervention

Yuxin Chen, Chen Tang, Jianglan Wei, Chenran Li, Ran Tian, Xiang Zhang, Wei Zhan, Peter Stone, Masayoshi Tomizuka

专题命中模仿学习与强化学习：embodied AI(abstract)；分类 cs.RO、cs.AI、cs.LG

2510.07730 2025-10-10 cs.LG cs.AI cs.RO 67%

DEAS: DEtached value learning with Action Sequence for Scalable Offline RL

Changyeon Kim, Haeone Lee, Younggyo Seo, Kimin Lee, Yuke Zhu

机构 * KAIST（韩国科学技术院）； UC Berkeley（伯克利大学）； University of Texas at Austin（德克萨斯大学奥斯汀分校）； NVIDIA（英伟达）

专题命中模仿学习与强化学习：manipulation(abstract)；分类 cs.RO、cs.AI、cs.LG

Comments Project website: https://changyeon.site/deas

2510.05619 2025-10-08 eess.AS 67%

Teaching Machines to Speak Using Articulatory Control

Akshay Anand, Chenxu Guo, Cheol Jun Cho, Jiachen Lian, Gopala Anumanchipalli

专题命中模仿学习与强化学习：manipulation(abstract)；robotic(abstract)

2509.15333 2025-09-22 cs.CV cs.AI cs.LG eess.IV 67%

Emulating Human-like Adaptive Vision for Efficient and Flexible Machine Visual Perception

Yulin Wang, Yang Yue, Yang Yue, Huanqian Wang, Haojun Jiang, Yizeng Han, Zanlin Ni, Yifan Pu, Minglei Shi, Rui Lu, Qisen Yang, Andrew Zhao, Zhuofan Xia, Shiji Song, Gao Huang

机构 * Learning And Perception (LEAP) Lab, Department of Automation, Tsinghua University（学习与感知（LEAP）实验室，自动化系，清华大学）

专题命中模仿学习与强化学习：embodied AI(abstract)；分类 cs.AI、cs.CV、cs.LG

2509.12531 2025-09-17 cs.RO cs.AI cs.LG cs.SY eess.SY 67%

Pre-trained Visual Representations Generalize Where it Matters in Model-Based Reinforcement Learning

Scott Jones, Liyou Zhou, Sebastian W. Pattinson

机构 * Institute for Manufacturing, Department of Engineering, University of Cambridge（剑桥大学制造研究所、工程系）

专题命中模仿学习与强化学习：robotic(abstract)；分类 cs.RO、cs.AI、cs.LG

2509.10423 2025-09-15 cs.AI cs.LG cs.RO 67%

Mutual Information Tracks Policy Coherence in Reinforcement Learning

Cameron Reid, Wael Hafez, Amirhossein Nazeri

专题命中模仿学习与强化学习：robotic(abstract)；分类 cs.RO、cs.AI、cs.LG

Comments 10 pages, 4 figures, 1 table

2505.22626 2025-09-10 cs.RO cs.AI cs.LG 67%

SCIZOR: A Self-Supervised Approach to Data Curation for Large-Scale Imitation Learning

Yu Zhang, Yuqi Xie, Huihan Liu, Rutav Shah, Michael Wan, Linxi Fan, Yuke Zhu

机构 * The University of Texas at Austin（德克萨斯大学奥斯汀分校）； NVIDIA Research（NVIDIA研究）

专题命中模仿学习与强化学习：robotic(abstract)；分类 cs.RO、cs.AI、cs.LG

2401.12497 2025-08-19 cs.AI cs.LG cs.RO 67%

Building Minimal and Reusable Causal State Abstractions for Reinforcement Learning

Zizhao Wang, Caroline Wang, Xuesu Xiao, Yuke Zhu, Peter Stone

专题命中模仿学习与强化学习：manipulation(abstract)；分类 cs.RO、cs.AI、cs.LG

Comments Accepted at AAAI24

Journal ref Proceedings of the Thirty-Eighth AAAI Conference on Artificial Intelligence (AAAI 2024), Article 1759, Pages 15778 - 15786

2506.05968 2025-08-14 cs.LG cs.AI cs.RO 67%

Gradual Transition from Bellman Optimality Operator to Bellman Operator in Online Reinforcement Learning

Motoki Omura, Kazuki Ota, Takayuki Osa, Yusuke Mukuta, Tatsuya Harada

机构 * The University of Tokyo（东京大学）

专题命中模仿学习与强化学习：manipulation(abstract)；分类 cs.RO、cs.AI、cs.LG

Comments Accepted at ICML 2025. Source code: https://github.com/motokiomura/annealed-q-learning

2508.05310 2025-08-08 cs.LG cs.AI cs.HC cs.RO 67%

ASkDAgger: Active Skill-level Data Aggregation for Interactive Imitation Learning

Jelle Luijkx, Zlatan Ajanović, Laura Ferranti, Jens Kober

机构 * Department of Cognitive Robotics（认知机器人学系）； Delft University of Technology（代尔夫特理工大学）； Department of Computer Science（计算机科学系）； RWTH Aachen University（亚琛工业大学）

专题命中模仿学习与强化学习：manipulation(abstract)；分类 cs.RO、cs.AI、cs.LG

Comments Accepted for publication in Transactions on Machine Learning Research (TMLR, 2025)

2411.05273 2025-08-07 cs.RO cs.AI cs.LG 67%

Real-World Offline Reinforcement Learning from Vision Language Model Feedback

Sreyas Venkataraman, Yufei Wang, Ziyu Wang, Navin Sriram Ravie, Zackory Erickson, David Held

机构 * Indian Institute of Technology, Kharagpur（印度理工学院，克哈格浦尔分校）； IIIS, Tsinghua University（清华大学人工智能研究所）； Robotics Institute, Carnegie Mellon University（卡内基梅隆大学机器人研究所）

专题命中模仿学习与强化学习：manipulation(abstract)；分类 cs.RO、cs.AI、cs.LG

Comments 7 pages. Accepted at the LangRob Workshop 2024 @ CoRL, 2024. Accepted at 2025 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2025)

2507.10843 2025-07-16 cs.LG cs.AI cs.RO 67%

Offline Reinforcement Learning with Wasserstein Regularization via Optimal Transport Maps

Motoki Omura, Yusuke Mukuta, Kazuki Ota, Takayuki Osa, Tatsuya Harada

专题命中模仿学习与强化学习：robotics(abstract)；分类 cs.RO、cs.AI、cs.LG

Comments Accepted at RLC 2025

2407.02508 2025-06-26 cs.RO cs.AI cs.LG 67%

Physics-informed Imitative Reinforcement Learning for Real-world Driving

Hang Zhou, Yihao Qin, Dan Xu, Yiding Ji

机构 * Robotics and Autonomous Systems Thrust, Systems Hub, The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州）机器人与自主系统 thrust、系统枢纽）； Department of Computer Science and Engineering, School of Engineering, The Hong Kong University of Science and Technology（香港科技大学计算机科学与工程系、工程学院）

专题命中模仿学习与强化学习：world model(abstract)；分类 cs.RO、cs.AI、cs.LG

2505.22642 2025-06-03 cs.RO cs.AI cs.LG 67%

FastTD3: Simple, Fast, and Capable Reinforcement Learning for Humanoid Control

Younggyo Seo, Carmelo Sferrazza, Haoran Geng, Michal Nauman, Zhao-Heng Yin, Pieter Abbeel

机构 * University of California, Berkeley（加州大学伯克利分校）； University of Warsaw（华沙大学）

专题命中模仿学习与强化学习：robotics(abstract)；分类 cs.RO、cs.AI、cs.LG

Comments Project webpage: https://younggyo.me/fast_td3

2501.17161 2025-05-27 cs.AI cs.CV cs.LG 67%

SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-training

Tianzhe Chu, Yuexiang Zhai, Jihan Yang, Shengbang Tong, Saining Xie, Dale Schuurmans, Quoc V. Le, Sergey Levine, Yi Ma

专题命中模仿学习与强化学习：navigation(abstract)；分类 cs.AI、cs.CV、cs.LG

Comments Website at https://tianzhechu.com/SFTvsRL

2505.09561 2025-05-21 cs.RO cs.AI cs.LG 67%

Learning Long-Context Diffusion Policies via Past-Token Prediction

Marcel Torne, Andy Tang, Yuejiang Liu, Chelsea Finn

机构 * Stanford University（斯坦福大学）

专题命中模仿学习与强化学习：robotic(abstract)；分类 cs.RO、cs.AI、cs.LG

Comments Videos are available at https://long-context-dp.github.io

2505.06883 2025-05-20 cs.RO cs.AI cs.LG 67%

FACET: Force-Adaptive Control via Impedance Reference Tracking for Legged Robots

Botian Xu, Haoyang Weng, Qingzhou Lu, Yang Gao, Huazhe Xu

机构 * Tsinghua University（清华大学）； Shanghai Qizhi Institute（上海启芝研究院）； Shanghai AI Lab（上海人工智能实验室）

专题命中模仿学习与强化学习：manipulation(abstract)；分类 cs.RO、cs.AI、cs.LG

2410.08893 2025-05-19 cs.LG cs.AI cs.RO 67%

Drama: Mamba-Enabled Model-Based Reinforcement Learning Is Sample and Parameter Efficient

Wenlong Wang, Ivana Dusparic, Yucheng Shi, Ke Zhang, Vinny Cahill

机构 * School of Computer Science and Statistics（计算机科学与统计学系）

专题命中模仿学习与强化学习：world model(abstract)；分类 cs.RO、cs.AI、cs.LG

Comments Published as a conference paper at ICLR 2025

2410.08852 2025-05-01 cs.RO cs.AI cs.HC cs.LG 67%

Conformalized Interactive Imitation Learning: Handling Expert Shift and Intermittent Feedback

Michelle Zhao, Reid Simmons, Henny Admoni, Aaditya Ramdas, Andrea Bajcsy

机构 * Robotics Institute, School of Computer Science, Carnegie Mellon University（机器人研究所、计算机科学学院、卡内基梅隆大学）； Departments of Statistics and Machine Learning, Carnegie Mellon University（统计学与机器学习系、卡内基梅隆大学）

专题命中模仿学习与强化学习：robotic(abstract)；分类 cs.RO、cs.AI、cs.LG

2406.08472 2025-04-22 cs.LG cs.AI cs.RO 67%

RILe: Reinforced Imitation Learning

Mert Albaba, Sammy Christen, Thomas Langarek, Christoph Gebhardt, Otmar Hilliges, Michael J. Black

机构 * ETH Zürich（苏黎世联邦理工学院）； Max Planck Institute for Intelligent Systems（智能系统马克斯·普朗克研究所）

专题命中模仿学习与强化学习：robotic(abstract)；分类 cs.RO、cs.AI、cs.LG

1. 模仿学习与强化学习 4113 篇

Vision-Proprioception Fusion with Mamba2 in End-to-End Reinforcement Learning for Motion Control

Online Action-Stacking Improves Reinforcement Learning Performance for Air Traffic Control

Flattening Hierarchies with Policy Bootstrapping

DHP: Discrete Hierarchical Planning for Hierarchical Reinforcement Learning Agents

Sample-Efficient Expert Query Control in Active Imitation Learning via Conformal Prediction

Green Resilience of Cyber-Physical Systems: Doctoral Dissertation

Are LLMs The Way Forward? A Case Study on LLM-Guided Reinforcement Learning for Decentralized Autonomous Driving

DeepSeek-Inspired Exploration of RL-based LLMs and Synergy with Wireless Networks: A Survey

Evolutionary Policy Optimization

Uncertainty-Based Smooth Policy Regularisation for Reinforcement Learning with Few Demonstrations

MEReQ: Max-Ent Residual-Q Inverse RL for Sample-Efficient Alignment from Intervention

DEAS: DEtached value learning with Action Sequence for Scalable Offline RL

Teaching Machines to Speak Using Articulatory Control

Emulating Human-like Adaptive Vision for Efficient and Flexible Machine Visual Perception

Pre-trained Visual Representations Generalize Where it Matters in Model-Based Reinforcement Learning

Mutual Information Tracks Policy Coherence in Reinforcement Learning

SCIZOR: A Self-Supervised Approach to Data Curation for Large-Scale Imitation Learning

Building Minimal and Reusable Causal State Abstractions for Reinforcement Learning

Gradual Transition from Bellman Optimality Operator to Bellman Operator in Online Reinforcement Learning

ASkDAgger: Active Skill-level Data Aggregation for Interactive Imitation Learning

Real-World Offline Reinforcement Learning from Vision Language Model Feedback

Offline Reinforcement Learning with Wasserstein Regularization via Optimal Transport Maps

Physics-informed Imitative Reinforcement Learning for Real-world Driving

FastTD3: Simple, Fast, and Capable Reinforcement Learning for Humanoid Control

SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-training

Learning Long-Context Diffusion Policies via Past-Token Prediction

FACET: Force-Adaptive Control via Impedance Reference Tracking for Legged Robots

Drama: Mamba-Enabled Model-Based Reinforcement Learning Is Sample and Parameter Efficient

Conformalized Interactive Imitation Learning: Handling Expert Shift and Intermittent Feedback

RILe: Reinforced Imitation Learning