世界模型 - arXivDaily 专题

2510.08226 2025-12-19 cs.LG 50%

UAMDP: Uncertainty-Aware Markov Decision Process for Risk-Constrained Reinforcement Learning from Probabilistic Forecasts

UAMDP：面向风险约束强化学习的概率预测不确定性感知马尔可夫决策过程

Michal Koren, Or Peretz, Tai Dinh, Philip S. Yu

机构 * The Kyoto College of Graduate Studies for Informatics（京都大学研究生院信息学研究科）； Department of Computer Science, University of Illinois at Chicago（伊利诺伊大学芝加哥分校计算机科学系）

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

AI总结 UAMDP通过整合概率预测、后验不确定性探索和风险约束规划，提升高波动环境下序列决策的安全性和盈利能力。

2512.15622 2025-12-18 eess.SY cs.SY 50%

Operator-Theoretic Joint Estimation of Aging-Aware State of Charge and Control-Informed State of Health

算子理论联合估计考虑老化因素的状态电量和基于控制信息的状态健康

Rahmat K. Adesunkanmi, Adel Alaeddini, Mahesh Krishnamurthy

专题命中模型式强化学习：latent dynamics(abstract)；dynamics model(abstract)

AI总结本文提出基于算子理论的联合估计框架，用于考虑老化因素的状态电量和基于控制信息的状态健康估计，通过端到端训练实现稳定且准确的容量预测。

Comments This paper has been submitted to IEEE for publication

2509.02196 2025-12-01 q-bio.BM cs.AI 50%

Beyond Ensembles: Simulating All-Atom Protein Dynamics in a Learned Latent Space

超越集成：在学习的潜在空间中模拟全部原子蛋白质动力学

Aditya Sengar, Jiying Zhang, Pierre Vandergheynst, Patrick Barth

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

AI总结本文提出GLDP模块，通过比较三种传播器，展示自回归神经网络在模拟全部原子蛋白质动力学中的优越性。

2511.20612 2025-11-26 cs.LG cs.SY eess.SY 50%

Sparse-to-Field Reconstruction via Stochastic Neural Dynamic Mode Decomposition

通过随机神经动态模式分解实现稀疏到场的重建

Yujin Kim, Sarah Dean

机构 * Cornell University（康奈尔大学）

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

AI总结本文提出随机NODE-DMD方法，通过建模连续时间非线性动态实现稀疏到场的重建，并量化预测不确定性，提升科学机器学习中动态系统建模的准确性与可靠性。

2103.08850 2025-11-21 cs.RO 50%

Few Shot System Identification for Reinforcement Learning

少样本系统识别用于强化学习

Karim Farid, Nourhan Sakr

机构 * Computer Science and Engineering Department（计算机科学与工程系）

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

AI总结本文提出了一种框架，通过学习动态条件于观测数据的概率分布，以提高强化学习中系统识别的鲁棒性和样本效率。

2501.08423 2025-11-20 stat.ML cs.LG 50%

LiLaN: A Linear Latent Network as the Solution Operator for Real-Time Solutions to Stiff Nonlinear Ordinary Differential Equations

William Cole Nockolds, C. G. Krishnanunni, Tan Bui-Thanh, Xianxhu Tang

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

2511.06854 2025-11-18 cs.LG stat.ML 50%

Beyond Observations: Reconstruction Error-Guided Irregularly Sampled Time Series Representation Learning

Jiexi Liu, Meng Cao, Songcan Chen

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

Comments Accepted by AAAI 2026

2511.04534 2025-11-18 cs.LG physics.ao-ph physics.comp-ph 50%

Uncertainty Quantification for Reduced-Order Surrogate Models Applied to Cloud Microphysics

Jonas E. Katona, Emily K. de Jong, Nipun Gunawardena

机构 * Department of Applied & Computational Mathematics（应用与计算数学系）； Yale University（耶鲁大学）； Atmospheric, Earth, & Energy Division（大气、地球与能源部门）； Lawrence Livermore National Laboratory（劳伦斯利弗莫尔国家实验室）

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

Comments Accepted at the NeurIPS 2025 Workshop on Machine Learning and the Physical Sciences (ML4PS). 12 pages, 3 figures, 2 tables. LLNL-CONF-2010541

2510.25674 2025-10-30 cs.LG 50%

Mechanistic Interpretability of RNNs emulating Hidden Markov Models

Elia Torre, Michele Viscione, Lucas Pompe, Benjamin F Grewe, Valerio Mante

机构 * Institute of Neuroinformatics, University of Zurich & ETH Zurich（神经信息学研究所，苏黎世大学及苏黎世联邦理工学院）

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

Comments Accepted at NeurIPS 2025

2312.02804 2025-10-30 cs.LG cs.PF math.OC math.PR 50%

Score-Aware Policy-Gradient and Performance Guarantees using Local Lyapunov Stability

Céline Comte, Matthieu Jonckheere, Jaron Sanders, Albert Senen-Cerda

机构 * LAAS–CNRS, Université de Toulouse, CNRS（LAAS–CNRS，图卢兹大学，CNRS）； Eindhoven University of Technology（埃因霍温理工大学）； LAAS-CNRS, IRIT, and Université de Toulouse（LAAS-CNRS，IRIT，图卢兹大学）

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

Journal ref Journal of Machine Learning Research 26, no. 132 (2025): 1-74

2505.14595 2025-10-28 cs.LG 50%

Physics-informed Reduced Order Modeling of Time-dependent PDEs via Differentiable Solvers

Nima Hosseini Dashtbayaz, Hesam Salehipour, Adrian Butscher, Nigel Morris

机构 * Department of Computer Science University of Western Ontario（计算机科学系大学西部安大略）

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

Comments Accepted at NeurIPS 2025

2510.19531 2025-10-23 cs.LG 50%

The Confusing Instance Principle for Online Linear Quadratic Control

Waris Radji, Odalric-Ambrym Maillard

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

Journal ref Reinforcement Learning Journal, 2025, 6

2510.18687 2025-10-22 cs.LG 50%

Reinforcement Learning with Imperfect Transition Predictions: A Bellman-Jensen Approach

Chenbei Lu, Zaiwei Chen, Tongxin Li, Chenye Wu, Adam Wierman

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

2510.18310 2025-10-22 cs.LG stat.ME 50%

Towards Identifiability of Hierarchical Temporal Causal Representation Learning

Zijian Li, Minghao Fu, Junxian Huang, Yifan Shen, Ruichu Cai, Yuewen Sun, Guangyi Chen, Kun Zhang

机构 * Carnegie Mellon University（卡内基梅隆大学）； Mohamed bin Zayed University of Artificial Intelligence（穆罕默德·本·扎耶德人工智能大学）； Guangdong University of Technology（广东技术大学）； University of California San Diego（加州大学圣地亚哥分校）

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

2510.07624 2025-10-14 stat.ML cs.LG 50%

From Data to Rewards: a Bilevel Optimization Perspective on Maximum Likelihood Estimation

Abdelhakim Benechehab, Gabriel Singer, Corentin Léger, Youssef Attia El Hili, Giuseppe Paolo, Albert Thomas, Maurizio Filippone, Balázs Kégl

机构 * Huawei Noah’s Ark Lab（华为诺亚实验室）； Department of Data Science, EURECOM（EURECOM数据科学系）； Cognizant AI Lab（Cognizant人工智能实验室）； Statistics Program, KAUST（KAUST统计学项目）

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

2510.03535 2025-10-07 cs.LG cs.NA math.NA stat.ML 50%

Sequential decoder training for improved latent space dynamics identification

William Anderson, Seung Whan Chung, Youngsoo Choi

机构 * Center for Applied Scientific Computing（应用科学计算中心）； Lawrence Livermore National Laboratory（劳伦斯利弗莫尔国家实验室）

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

2510.03207 2025-10-06 cs.LG 50%

To Distill or Decide? Understanding the Algorithmic Trade-off in Partially Observable Reinforcement Learning

Yuda Song, Dhruv Rohatgi, Aarti Singh, J. Andrew Bagnell

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

Comments 45 pages, 9 figures, published at NeurIPS 2025

2509.12149 2025-09-16 physics.flu-dyn 50%

Superresolving Non-linear PDE Dynamics with Reduced-Order Autodifferentiable Ensemble Kalman Filtering For Turbulence Modeling and Flow Regulation

Mrigank Dhingra, Omer San

专题命中模型式强化学习：latent dynamics(abstract)；dynamics model(abstract)

2405.14425 2025-08-26 cs.LG stat.ML 50%

When predict can also explain: few-shot prediction to select better neural latents

Kabir Dabholkar, Omri Barak

机构 * Faculty of Mathematics, Technion – Israel Institute of Technology（数学系，技术离子理工学院）； Rappaport Faculty of Medicine and Network Biology Research Laboratory, Technion - Israel Institute of Technology（医学与网络生物学研究实验室，技术离子理工学院）

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

2508.14295 2025-08-21 cs.CV cs.AI cs.LG 50%

Pixels to Play: A Foundation Model for 3D Gameplay

Yuguang Yue, Chris Green, Samuel Hunt, Irakli Salia, Wenzhe Shi, Jonathan J Hunt

机构 * Player2

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

Journal ref Conference on Games 2025 (Short paper)

2405.02754 2025-08-15 cs.RO cs.AI cs.LG 50%

Implicit Safe Set Algorithm for Provably Safe Reinforcement Learning

Weiye Zhao, Feihan Li, Changliu Liu

机构 * Carnegie Mellon University（卡内基梅隆大学）

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

Comments Accepted to Journal of Artificial Intelligence Research. arXiv admin note: text overlap with arXiv:2308.13140

2406.14801 2025-08-12 q-bio.NC cs.SY eess.SY q-bio.QM 50%

Model Predictive Control on the Neural Manifold

Christof Fehrman, C. Daniel Meliza

专题命中模型式强化学习：latent dynamics(abstract)；dynamics model(abstract)

2507.12494 2025-07-18 cs.AI cs.GT cs.MA cs.RO 50%

MR-LDM -- The Merge-Reactive Longitudinal Decision Model: Game Theoretic Human Decision Modeling for Interactive Sim Agents

Dustin Holley, Jovin D'sa, Hossein Nourkhiz Mahjoub, Gibran Ali

机构 * Global Center for Automotive Performance Simulation（全球汽车性能模拟中心）； Honda Research Institute, USA Inc.（本田研究院（美国公司））； Division of Data and Analytics, Virginia Tech Transportation Institute（弗吉尼亚理工运输研究所数据分析部门）

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

Comments 8 pages

2506.12358 2025-06-17 cs.LG cs.SY eess.SY 50%

Relative Entropy Regularized Reinforcement Learning for Efficient Encrypted Policy Synthesis

Jihoon Suh, Yeongjun Jang, Kaoru Teranishi, Takashi Tanaka

机构 * School of Aeronautics and Astronautics, Purdue University（航空宇航工程学院，普渡大学）； ASRI, the Department of Electrical and Computer Engineering, Seoul National University（ASRI，电气电子工程系，首尔国立大学）； Japan Society for the Promotion of Science（日本学术振兴会）

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

Comments 6 pages, 2 figures, Published in IEEE Control Systems Letters, June 2025

Journal ref IEEE Control Systems Letters, pp. 1-1, June 2025

2501.08950 2025-06-16 cs.LO cs.LG 50%

Approximating Fixpoints of Approximated Functions

Paolo Baldan, Sebastian Gurke, Barbara König, Tommaso Padoan, Florian Wittbold

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

2506.04399 2025-06-06 cs.LG cs.AI cs.RO 50%

Unsupervised Meta-Testing with Conditional Neural Processes for Hybrid Meta-Reinforcement Learning

Suzan Ece Ada, Emre Ugur

机构 * Department of Computer Engineering, Bogazici University（计算机工程系，博兹达奇大学）

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

Comments Published in IEEE Robotics and Automation Letters Volume: 9, Issue: 10, 8427 - 8434, October 2024. 8 pages, 7 figures

Journal ref IEEE Robotics and Automation Letters Volume: 9, Issue: 10, 8427 - 8434, October 2024,

2412.11743 2025-06-03 stat.ML cs.LG stat.ME 50%

Generalized Bayesian deep reinforcement learning

Shreya Sinha Roy, Richard G. Everitt, Christian P. Robert, Ritabrata Dutta

机构 * University of Warwick（沃里克大学）； CEREMADE, Université Paris Dauphine PSL（巴黎第纳大学CEREMADE）

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

2410.00665 2025-06-02 q-bio.NC cs.LG cs.NE 50%

Graph-Based Representation Learning of Neuronal Dynamics and Behavior

Moein Khajehnejad, Forough Habibollahi, Ahmad Khajehnejad, Chris French, Brett J. Kagan, Adeel Razi

机构 * Turner Institute for Brain and Mental Health（大脑与心理健康研究所）； School of Psychological Sciences, Monash University（墨尔本大学心理学科学学院）； Cortical Labs（皮层实验室）

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

Comments 31 pages, 6 figures, 4 supplemental figures, 4 tables, 8 supplemental tables

2502.19459 2025-05-28 cs.CV cs.GR cs.LG cs.RO 50%

ArtGS: Building Interactable Replicas of Complex Articulated Objects via Gaussian Splatting

Yu Liu, Baoxiong Jia, Ruijie Lu, Junfeng Ni, Song-Chun Zhu, Siyuan Huang

机构 * Tsinghua University（清华大学）； State Key Laboratory of General Artificial Intelligence, BIGAI（通用人工智能国家重点实验室）； Peking University（北京大学）

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

2503.18787 2025-05-14 cs.LG math.OC 50%

Sample-Efficient Reinforcement Learning of Koopman eNMPC

Daniel Mayfrank, Mehmet Velioglu, Alexander Mitsos, Manuel Dahmen

机构 * Forschungszentrum Jülich GmbH, Institute of Climate and Energy Systems, Energy Systems Engineering (ICE-1)（茹里希研究中心有限公司，气候与能源系统研究所，能源系统工程（ICE-1））； RWTH Aachen University（亚琛工业大学）； JARA-ENERGY ； RWTH Aachen University, Process Systems Engineering (AVT.SVT)（亚琛工业大学，过程系统工程（AVT.SVT））

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

Comments 25 pages, 9 figures, 2 tables

1. 模型式强化学习 1124 篇

UAMDP: Uncertainty-Aware Markov Decision Process for Risk-Constrained Reinforcement Learning from Probabilistic Forecasts

Operator-Theoretic Joint Estimation of Aging-Aware State of Charge and Control-Informed State of Health

Beyond Ensembles: Simulating All-Atom Protein Dynamics in a Learned Latent Space

Sparse-to-Field Reconstruction via Stochastic Neural Dynamic Mode Decomposition

Few Shot System Identification for Reinforcement Learning

LiLaN: A Linear Latent Network as the Solution Operator for Real-Time Solutions to Stiff Nonlinear Ordinary Differential Equations

Beyond Observations: Reconstruction Error-Guided Irregularly Sampled Time Series Representation Learning

Uncertainty Quantification for Reduced-Order Surrogate Models Applied to Cloud Microphysics

Mechanistic Interpretability of RNNs emulating Hidden Markov Models

Score-Aware Policy-Gradient and Performance Guarantees using Local Lyapunov Stability

Physics-informed Reduced Order Modeling of Time-dependent PDEs via Differentiable Solvers

The Confusing Instance Principle for Online Linear Quadratic Control

Reinforcement Learning with Imperfect Transition Predictions: A Bellman-Jensen Approach

Towards Identifiability of Hierarchical Temporal Causal Representation Learning

From Data to Rewards: a Bilevel Optimization Perspective on Maximum Likelihood Estimation

Sequential decoder training for improved latent space dynamics identification

To Distill or Decide? Understanding the Algorithmic Trade-off in Partially Observable Reinforcement Learning

Superresolving Non-linear PDE Dynamics with Reduced-Order Autodifferentiable Ensemble Kalman Filtering For Turbulence Modeling and Flow Regulation

When predict can also explain: few-shot prediction to select better neural latents

Pixels to Play: A Foundation Model for 3D Gameplay

Implicit Safe Set Algorithm for Provably Safe Reinforcement Learning

Model Predictive Control on the Neural Manifold

MR-LDM -- The Merge-Reactive Longitudinal Decision Model: Game Theoretic Human Decision Modeling for Interactive Sim Agents

Relative Entropy Regularized Reinforcement Learning for Efficient Encrypted Policy Synthesis

Approximating Fixpoints of Approximated Functions

Unsupervised Meta-Testing with Conditional Neural Processes for Hybrid Meta-Reinforcement Learning

Generalized Bayesian deep reinforcement learning

Graph-Based Representation Learning of Neuronal Dynamics and Behavior

ArtGS: Building Interactable Replicas of Complex Articulated Objects via Gaussian Splatting

Sample-Efficient Reinforcement Learning of Koopman eNMPC