世界模型 - arXivDaily 专题

2106.09119 2021-06-21 cs.LG 65%

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

2101.07156 2021-04-16 eess.SY cs.SY 65%

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

2009.04278 2020-09-10 cs.LG cs.SY eess.SY stat.ML 65%

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

2608.16431 2026-08-18 eess.SY cs.SY 新提交 64%

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)

1902.08705 2026-06-04 cs.RO cs.AI cs.LG cs.SY eess.SY 64%

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）

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/

2605.16692 2026-05-20 cs.LG cs.AI cs.RO 64%

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

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。

2604.19980 2026-04-23 cs.RO cs.SY eess.SY 64%

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）

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。

2212.14511 2026-03-10 cs.LG cs.SY eess.SY math.OC stat.ML 64%

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）

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

2412.14312 2025-06-24 cs.LG 64%

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）

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

2402.05421 2025-06-16 cs.LG cs.AI cs.RO 64%

DiffTORI: Differentiable Trajectory Optimization for Deep Reinforcement and Imitation Learning

Weikang Wan, Ziyu Wang, Yufei Wang, Zackory Erickson, David Held

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

Comments NeurIPS 2024 (Spotlight)

2503.11964 2025-05-30 cs.LG stat.ML 64%

Entropy-regularized Gradient Estimators for Approximate Bayesian Inference

Jasmeet Kaur

机构 * Department of Computer Science University of Texas, Austin（计算机科学系德克萨斯大学奥斯汀分校）

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

2503.17693 2025-03-25 cs.NI 64%

Conditional Diffusion Model with OOD Mitigation as High-Dimensional Offline Resource Allocation Planner in Clustered Ad Hoc Networks

Kechen Meng, Sinuo Zhang, Rongpeng Li, Chan Wang, Ming Lei, Zhifeng Zhao

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

Comments This work has been submitted to the IEEE for possible publication

2410.09972 2024-10-15 cs.LG cs.AI cs.CV cs.RO 64%

Make the Pertinent Salient: Task-Relevant Reconstruction for Visual Control with Distractions

Kyungmin Kim, JB Lanier, Pierre Baldi, Charless Fowlkes, Roy Fox

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

2209.10200 2023-07-12 cs.LG 64%

Performance Optimization for Variable Bitwidth Federated Learning in Wireless Networks

Sihua Wang, Mingzhe Chen, Christopher G. Brinton, Changchuan Yin, Walid Saad, Shuguang Cui

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

2301.11520 2023-06-01 cs.LG cs.AI cs.CV cs.RO 64%

SNeRL: Semantic-aware Neural Radiance Fields for Reinforcement Learning

Dongseok Shim, Seungjae Lee, H. Jin Kim

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

Comments ICML 2023. First two authors contributed equally. Order was determined by coin flip

2011.07401 2022-04-08 cs.PF cs.LG 64%

RL-QN: A Reinforcement Learning Framework for Optimal Control of Queueing Systems

Bai Liu, Qiaomin Xie, Eytan Modiano

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

2106.14421 2021-06-29 cs.LG 64%

Causal Reinforcement Learning using Observational and Interventional Data

Maxime Gasse, Damien Grasset, Guillaume Gaudron, Pierre-Yves Oudeyer

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

2008.12775 2021-05-28 cs.LG cs.AI cs.RO stat.ML 64%

On the model-based stochastic value gradient for continuous reinforcement learning

Brandon Amos, Samuel Stanton, Denis Yarats, Andrew Gordon Wilson

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)

Comments L4DC 2021

2010.13957 2021-01-12 cs.LG cs.AI cs.CV cs.RO 64%

MELD: Meta-Reinforcement Learning from Images via Latent State Models

Tony Z. Zhao, Anusha Nagabandi, Kate Rakelly, Chelsea Finn, Sergey Levine

专题命中模型式强化学习：latent dynamics(abstract)；分类 cs.AI、cs.LG、cs.CV

Comments Accepted to CoRL 2020. Supplementary material at https://sites.google.com/view/meld-lsm/home . 16 pages, 19 figures. V2: add funding acknowledgements, reduce file size

2010.04893 2020-10-13 cs.LG 64%

Trust the Model When It Is Confident: Masked Model-based Actor-Critic

Feiyang Pan, Jia He, Dandan Tu, Qing He

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 NeurIPS 2020

2004.08763 2020-04-21 cs.LG cs.AI cs.RO stat.ML 64%

Model-Predictive Control via Cross-Entropy and Gradient-Based Optimization

Homanga Bharadhwaj, Kevin Xie, Florian Shkurti

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)

Comments L4DC 2020; Accepted for presentation in the 2nd Annual Conference on Learning for Dynamics and Control

1911.03594 2019-11-21 cs.LG cs.AI cs.RO stat.ML 64%

Robo-PlaNet: Learning to Poke in a Day

Maxime Chevalier-Boisvert, Guillaume Alain, Florian Golemo, Derek Nowrouzezahrai

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)

Comments 4 pages, 3 figures. Version 2: added reference and acknowledgement

1902.00673 2019-02-05 cs.AI 64%

Belief dynamics extraction

Arun Kumar, Zhengwei Wu, Xaq Pitkow, Paul Schrater

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

1806.07371 2018-10-31 cs.CV cs.AI cs.LG 64%

Object-Oriented Dynamics Predictor

Guangxiang Zhu, Zhiao Huang, Chongjie Zhang

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

Comments Accepted to NIPS 2018

1708.02596 2017-12-05 cs.LG cs.AI cs.RO 64%

Neural Network Dynamics for Model-Based Deep Reinforcement Learning with Model-Free Fine-Tuning

Anusha Nagabandi, Gregory Kahn, Ronald S. Fearing, Sergey Levine

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)

1710.00489 2026-06-04 cs.RO cs.AI cs.CV cs.NE cs.SY eess.SY 62%

SE3-Pose-Nets: Structured Deep Dynamics Models for Visuomotor Planning and Control

SE3-姿态网络：用于视觉-运动规划和控制的结构深度动力学模型

Arunkumar Byravan, Felix Leeb, Franziska Meier, Dieter Fox

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

AI总结本文提出了一种基于结构深度动力学模型的深度视觉-运动控制方法，通过编码器-解码器结构学习低维姿态嵌入，实现场景分割和姿态预测，并在现实世界中实现了闭环控制。

Comments 8 pages, Initial submission to IEEE International Conference on Robotics and Automation (ICRA) 2018

2502.04892 2025-02-10 cs.LG q-bio.NC stat.ML 62%

A Foundational Brain Dynamics Model via Stochastic Optimal Control

Joonhyeong Park, Byoungwoo Park, Chang-Bae Bang, Jungwon Choi, Hyungjin Chung, Byung-Hoon Kim, Juho Lee

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

Comments The first two authors contributed equally

2312.04374 2024-12-03 cs.RO cs.AI cs.LG 62%

Deep Dynamics: Vehicle Dynamics Modeling with a Physics-Constrained Neural Network for Autonomous Racing

John Chrosniak, Jingyun Ning, Madhur Behl

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

Comments Published in the IEEE Robotics and Automation Letters and presented at the IEEE International Conference on Intelligent Robots and Systems

Journal ref IEEE Robotics and Automation Letters (Volume: 9, Issue: 6, June 2024), 5292 - 5297

2410.18912 2024-10-25 cs.RO cs.AI cs.LG 62%

Dynamic 3D Gaussian Tracking for Graph-Based Neural Dynamics Modeling

Mingtong Zhang, Kaifeng Zhang, Yunzhu Li

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

Comments Project Page: https://gs-dynamics.github.io

2407.01418 2024-07-02 cs.RO cs.AI cs.LG 62%

RoboPack: Learning Tactile-Informed Dynamics Models for Dense Packing

Bo Ai, Stephen Tian, Haochen Shi, Yixuan Wang, Cheston Tan, Yunzhu Li, Jiajun Wu

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

Comments Robotics: Science and Systems (RSS), 2024. Project page: https://robo-pack.github.io/

1. 模型式强化学习 1124 篇

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

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

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

Stable Multi-Step Rollouts via Uncertainty-Guided Hybrid Dynamics

A General Framework for Structured Learning of Mechanical Systems

EfficientTDMPC: Improved MPC Objectives for Sample-Efficient Continuous Control

Efficient Reinforcement Learning using Linear Koopman Dynamics for Nonlinear Robotic Systems

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

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

DiffTORI: Differentiable Trajectory Optimization for Deep Reinforcement and Imitation Learning

Entropy-regularized Gradient Estimators for Approximate Bayesian Inference

Conditional Diffusion Model with OOD Mitigation as High-Dimensional Offline Resource Allocation Planner in Clustered Ad Hoc Networks

Make the Pertinent Salient: Task-Relevant Reconstruction for Visual Control with Distractions

Performance Optimization for Variable Bitwidth Federated Learning in Wireless Networks

SNeRL: Semantic-aware Neural Radiance Fields for Reinforcement Learning

RL-QN: A Reinforcement Learning Framework for Optimal Control of Queueing Systems

Causal Reinforcement Learning using Observational and Interventional Data

On the model-based stochastic value gradient for continuous reinforcement learning

MELD: Meta-Reinforcement Learning from Images via Latent State Models

Trust the Model When It Is Confident: Masked Model-based Actor-Critic

Model-Predictive Control via Cross-Entropy and Gradient-Based Optimization

Robo-PlaNet: Learning to Poke in a Day

Belief dynamics extraction

Object-Oriented Dynamics Predictor

Neural Network Dynamics for Model-Based Deep Reinforcement Learning with Model-Free Fine-Tuning

SE3-Pose-Nets: Structured Deep Dynamics Models for Visuomotor Planning and Control

A Foundational Brain Dynamics Model via Stochastic Optimal Control

Deep Dynamics: Vehicle Dynamics Modeling with a Physics-Constrained Neural Network for Autonomous Racing

Dynamic 3D Gaussian Tracking for Graph-Based Neural Dynamics Modeling

RoboPack: Learning Tactile-Informed Dynamics Models for Dense Packing