世界模型

Why and How Auxiliary Tasks Improve JEPA Representations

Jiacan Yu, Siyi Chen, Mingrui Liu, Nono Horiuchi, Vladimir Braverman, Zicheng Xu, Dan Haramati, Randall Balestriero

机构 * Johns Hopkins University（约翰霍普金斯大学） ； Northwestern University（西北大学） ； University of Rochester（罗切斯特大学） ； Brown University（布朗大学） 纠错

专题命中 模型式强化学习 ：latent dynamics(abstract)；model-based RL(abstract)；分类 cs.AI、cs.LG

IQL-TD-MPC: Implicit Q-Learning for Hierarchical Model Predictive Control

Rohan Chitnis, Yingchen Xu, Bobak Hashemi, Lucas Lehnert, Urun Dogan, Zheqing Zhu, Olivier Delalleau

专题命中 模型式强化学习 ：model-based reinforcement learning(abstract)；model-based RL(abstract)；分类 cs.AI、cs.LG

Journal ref Short version published at ICRA 2024 (https://tinyurl.com/icra24-iqltdmpc)

Optimistic MLE -- A Generic Model-based Algorithm for Partially Observable Sequential Decision Making

Qinghua Liu, Praneeth Netrapalli, Csaba Szepesvári, Chi Jin

专题命中 模型式强化学习 ：model-based reinforcement learning(abstract)；model-based RL(abstract)；分类 cs.AI、cs.LG

Learning Parsimonious Dynamics for Generalization in Reinforcement Learning

Tankred Saanum, Eric Schulz

专题命中 模型式强化学习 ：latent dynamics(abstract)；model-based reinforcement learning(abstract)；分类 cs.LG；dynamics model(abstract)

Algorithmic Framework for Model-based Deep Reinforcement Learning with Theoretical Guarantees

Yuping Luo, Huazhe Xu, Yuanzhi Li, Yuandong Tian, Trevor Darrell, Tengyu Ma

专题命中 模型式强化学习 ：model-based reinforcement learning(abstract)；model-based RL(abstract)；分类 cs.AI、cs.LG

Comments Added important notes that the conditions of Theorem 3.1 cannot simultaneously hold for most models

The LoCA Regret: A Consistent Metric to Evaluate Model-Based Behavior in Reinforcement Learning

Harm van Seijen, Hadi Nekoei, Evan Racah, Sarath Chandar

专题命中 模型式强化学习 ：model-based reinforcement learning(abstract)；model-based RL(abstract)；分类 cs.AI、cs.LG

Comments NeurIPS 2020, code: https://github.com/chandar-lab/LoCA

Trajectory-wise Multiple Choice Learning for Dynamics Generalization in Reinforcement Learning

Younggyo Seo, Kimin Lee, Ignasi Clavera, Thanard Kurutach, Jinwoo Shin, Pieter Abbeel

专题命中 模型式强化学习 ：model-based reinforcement learning(abstract)；model-based RL(abstract)；分类 cs.LG；dynamics model(abstract)

Comments Accepted in NeurIPS2020. First two authors contributed equally, website: https://sites.google.com/view/trajectory-mcl code: https://github.com/younggyoseo/trajectory_mcl

Efficient Bayes-Adaptive Reinforcement Learning using Sample-Based Search

Arthur Guez, David Silver, Peter Dayan

专题命中 模型式强化学习 ：model-based reinforcement learning(abstract)；model-based RL(abstract)；分类 cs.AI、cs.LG

Comments 14 pages, 7 figures, includes supplementary material. Advances in Neural Information Processing Systems (NIPS) 2012

Journal ref (2012) Advances in Neural Information Processing Systems 25, pages 1034-1042

LNODE: latent dynamics reveal the shared spatiotemporal structure of amyloid-$β$ progression

LNODE：潜变量揭示阿尔茨海默病β淀粉样蛋白进展的共享时空结构

Zheyu Wen, George Biros

专题命中 模型式强化学习 ：latent dynamics(title)

AI总结 LNODE模型通过PET影像校准，揭示阿尔茨海默病β淀粉样蛋白进展的时空结构，具备融合、定量分析和解释能力，展现高参数可识别性和稳定性。

Comments 38 pages, 13 figures

Optimal Control of Probabilistic Dynamics Models via Mean Hamiltonian Minimization

David Leeftink, Çağatay Yıldız, Steffen Ridderbusch, Max Hinne, Marcel van Gerven

机构 * Department of Machine Learning and Neural Computing, Donders Institute for Brain, Cognition and Behaviour, Radboud University（机器学习与神经计算系，多纳尔斯脑认知行为研究所，拉德堡德大学） ； Cluster of Excellence Machine Learning, University of Tübingen（卓越机器学习集群，图宾根大学） ； Department of Engineering, Control Group, University of Oxford（工程系，控制组，牛津大学） 纠错

专题命中 模型式强化学习 ：dynamics model(title,abstract)；model-based reinforcement learning(abstract)；分类 cs.LG

Comments 8 pages, 4 figures, 2 tables

Probabilistically safe and efficient model-based reinforcement learning

Filippo Airaldi, Bart De Schutter, Azita Dabiri

专题命中 模型式强化学习 ：model-based reinforcement learning(title)

Comments 8 pages, 4 figures, accepted to 2025 CDC

Learning safety in model-based Reinforcement Learning using MPC and Gaussian Processes

Filippo Airaldi, Bart De Schutter, Azita Dabiri

专题命中 模型式强化学习 ：model-based reinforcement learning(title)

Comments 8 pages, 3 figures, submitted to IFAC World Congress 2023 (implemented reviews)

Journal ref IFAC-PapersOnLine, 56 (2), 2023, 5759-5764

Bimanual Deformable Bag Manipulation Using a Structure-of-Interest Based Neural Dynamics Model

Peng Zhou, Pai Zheng, Jiaming Qi, Chenxi Li, Samantha Lee, Chenguang Yang, David Navarro-Alarcon, Jia Pan

专题命中 模型式强化学习 ：dynamics model(title,abstract)；latent dynamics(abstract)；分类 cs.RO

Reinforcement Twinning: from digital twins to model-based reinforcement learning

Lorenzo Schena, Pedro Marques, Romain Poletti, Samuel Ahizi, Jan Van den Berghe, Miguel A. Mendez

专题命中 模型式强化学习 ：model-based reinforcement learning(title)

Comments submitted Journal of Computational Science

On learning latent dynamics of the AUG plasma state

A. Kit, A. E. Järvinen, Y. R. J. Poels, S. Wiesen, V. Menkovski, R. Fischer, M. Dunne, ASDEX-Upgrade Team

专题命中 模型式强化学习 ：latent dynamics(title)

Comments 9 pages, 11 figures, 4th International Conference on Data-Driven Plasma Sciences

Journal ref Phys. Plasmas 31, 032504 (2024)

Efficient Preference-Based Reinforcement Learning Using Learned Dynamics Models

Yi Liu, Gaurav Datta, Ellen Novoseller, Daniel S. Brown

专题命中 模型式强化学习 ：dynamics model(title,abstract)；model-based RL(abstract)；分类 cs.LG

Comments In proceedings of the 2023 IEEE International Conference on Robotics and Automation (ICRA 2023)

Dynamic Programming-based Approximate Optimal Control for Model-Based Reinforcement Learning

Prakash Mallick, Zhiyong Chen

专题命中 模型式强化学习 ：model-based reinforcement learning(title)

Comments 12 Pages, 6 figures. arXiv admin note: text overlap with arXiv:2010.00207

Reward-Consistent Dynamics Models are Strongly Generalizable for Offline Reinforcement Learning

Fan-Ming Luo, Tian Xu, Xingchen Cao, Yang Yu

专题命中 模型式强化学习 ：dynamics model(title,abstract)；model-based reinforcement learning(abstract)；分类 cs.LG

Neural-network learning of SPOD latent dynamics

Andrea Lario, Romit Maulik, Oliver T. Schmidt, Gianluigi Rozza, Gianmarco Mengaldo

专题命中 模型式强化学习 ：latent dynamics(title)

Comments 29 pages, 18 figures, 6 tables

TartanDrive: A Large-Scale Dataset for Learning Off-Road Dynamics Models

Samuel Triest, Matthew Sivaprakasam, Sean J. Wang, Wenshan Wang, Aaron M. Johnson, Sebastian Scherer

专题命中 模型式强化学习 ：dynamics model(title,abstract)；model-based reinforcement learning(abstract)；分类 cs.RO

Investigating Compounding Prediction Errors in Learned Dynamics Models

Nathan Lambert, Kristofer Pister, Roberto Calandra

专题命中 模型式强化学习 ：dynamics model(title,abstract)；model-based reinforcement learning(abstract)；分类 cs.LG

Comments 25 pages, 19 figures

Behavioral Priors and Dynamics Models: Improving Performance and Domain Transfer in Offline RL

Catherine Cang, Aravind Rajeswaran, Pieter Abbeel, Michael Laskin

专题命中 模型式强化学习 ：dynamics model(title,abstract)；model-based RL(abstract)；分类 cs.LG

Model-Based Reinforcement Learning for Approximate Optimal Control with Temporal Logic Specifications

Max Cohen, Calin Belta

专题命中 模型式强化学习 ：model-based reinforcement learning(title)

Comments To appear at the 24th ACM International Conference on Hybrid Systems: Computation and Control

DyNODE: Neural Ordinary Differential Equations for Dynamics Modeling in Continuous Control

Victor M. Martinez Alvarez, Rareş Roşca, Cristian G. Fălcuţescu

专题命中 模型式强化学习 ：dynamics model(title,abstract)；model-based reinforcement learning(abstract)；分类 cs.LG

Comments 9 pages, 5 figures

Stable Multi-Step Rollouts via Uncertainty-Guided Hybrid Dynamics

基于不确定性引导混合动力学的稳定多步回滚

Andrei Maalberg, Axel Neumann, Jens Knobloch

专题命中 模型式强化学习 ：model-based reinforcement learning(abstract)；model-based RL(abstract)；dynamics model(abstract)

AI总结 本文提出一种与模型无关的不确定性引导混合动力学框架，用于解决基于模型的强化学习中多步回滚不稳定的问题，在杜芬振子实验中实现了稳定长 horizon 预测并改善了成本-努力权衡。

Comments Accepted for presentation at, and publication in the Proceedings of the 65th IEEE Conference on Decision and Control (CDC 2026)

A General Framework for Structured Learning of Mechanical Systems

结构机械系统学习的通用框架

Jayesh K. Gupta, Kunal Menda, Zachary Manchester, Mykel J. Kochenderfer

机构 * Stanford University（斯坦福大学） 纠错

专题命中 模型式强化学习 ：model-based reinforcement learning(abstract)；分类 cs.AI、cs.LG、cs.RO；dynamics model(abstract)

AI总结 本文提出了一种通用框架，用于结构化学习机械系统，通过结合先验知识和训练表达式近似器来提高模型的准确性和效率。

Comments 10 pages, 7 figures. First two authors contributed equally. Submitted to IROS/RA-L. Code at https://github.com/sisl/mechamodlearn/

EfficientTDMPC: Improved MPC Objectives for Sample-Efficient Continuous Control

EfficientTDMPC: 改进的MPC目标以实现高效的连续控制

Thomas Evers, Cristian Meo, Wendelin Bohmer, Justin Dauwels, Yaniv Oren

机构 * TU Delft（代尔夫特理工大学） ； LatentWorlds AI 纠错

专题命中 模型式强化学习 ：model-based reinforcement learning(abstract)；分类 cs.AI、cs.LG、cs.RO；dynamics model(abstract)

AI总结 本文提出EfficientTDMPC，一种基于模型的强化学习方法，用于连续控制，通过减少误差和增加数据新鲜度来提高样本效率。

Efficient Reinforcement Learning using Linear Koopman Dynamics for Nonlinear Robotic Systems

利用线性Koopman动力学实现非线性机器人系统的高效强化学习

Wenjian Hao, Yuxuan Fang, Zehui Lu, Shaoshuai Mou

机构 * School of Aeronautics and Astronautics, Purdue University（航空航天学院，普渡大学） 纠错

专题命中 模型式强化学习 ：model-based reinforcement learning(abstract)；model-based RL(abstract)；分类 cs.RO

AI总结 本文提出基于模型的强化学习框架，通过Koopman算子理论学习非线性机器人系统的线性提升动力学，并将其整合到actor-critic架构中以优化策略，实验显示在模拟和现实平台中样本效率和控制性能均优于基线方法。

Cost-Driven Representation Learning for Linear Quadratic Gaussian Control: Part I

基于成本驱动的线性二次高斯控制的状态表示学习：第一部分

Yi Tian, Kaiqing Zhang, Russ Tedrake, Suvrit Sra

机构 * Massachusetts Institute of Technology（麻省理工学院） ； University of Maryland, College Park（马里兰大学 College Park 分校） ； Technical University Munich（慕尼黑技术大学） 纠错

专题命中 模型式强化学习 ：latent dynamics(abstract)；model-based reinforcement learning(abstract)；分类 cs.LG

AI总结 本文提出了一种基于成本驱动的方法，用于学习状态表示以解决线性二次高斯控制问题，并建立了有限样本下的保证。

Comments 51 pages; preliminary version appeared in L4DC 2023; this is the extended journal version, with an end-to-end guarantee added

Stealing That Free Lunch: Exposing the Limits of Dyna-Style Reinforcement Learning

Brett Barkley, David Fridovich-Keil

机构 * Department of Computer Science, University of Texas at Austin, Austin, TX, USA（德克萨斯大学奥斯汀分校计算机科学系） ； Department of Aerospace Engineering and Engineering Mechanics, University of Texas at Austin, Austin, TX, USA（德克萨斯大学奥斯汀分校航空航天工程与工程力学系） 纠错

专题命中 模型式强化学习 ：model-based reinforcement learning(abstract)；model-based RL(abstract)；分类 cs.LG

Comments Accepted to ICML 2025