世界模型 - arXivDaily 专题

2405.18193 2024-05-29 cs.LG cs.CV 94%

In-Context Symmetries: Self-Supervised Learning through Contextual World Models

Sharut Gupta, Chenyu Wang, Yifei Wang, Tommi Jaakkola, Stefanie Jegelka

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

Comments 32 pages, 24 tables and 11 figures

2405.15083 2024-05-27 cs.AI cs.CV 94%

MuDreamer: Learning Predictive World Models without Reconstruction

Maxime Burchi, Radu Timofte

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

2405.05890 2024-05-10 cs.LG cs.AI 94%

Safe Exploration Using Bayesian World Models and Log-Barrier Optimization

Yarden As, Bhavya Sukhija, Andreas Krause

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

2404.00462 2024-05-06 cs.LG cs.RO 94%

Zero-shot Safety Prediction for Autonomous Robots with Foundation World Models

Zhenjiang Mao, Siqi Dai, Yuang Geng, Ivan Ruchkin

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

Comments Presented at the Back to the Future-Robot Learning Going Probabilistic Workshop, co-located with ICRA 2024. https://openreview.net/forum?id=gHhBNIq9Cs

2312.08533 2024-03-28 cs.LG cs.AI 94%

World Models via Policy-Guided Trajectory Diffusion

Marc Rigter, Jun Yamada, Ingmar Posner

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

Comments Published in TMLR, March 2024

2403.07944 2024-03-14 cs.CV cs.AI 94%

WorldGPT: A Sora-Inspired Video AI Agent as Rich World Models from Text and Image Inputs

Deshun Yang, Luhui Hu, Yu Tian, Zihao Li, Chris Kelly, Bang Yang, Cindy Yang, Yuexian Zou

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

Comments 11 pages, 2 figures, 2 tables

2402.05290 2024-02-13 cs.LG cs.AI 94%

Do Transformer World Models Give Better Policy Gradients?

Michel Ma, Tianwei Ni, Clement Gehring, Pierluca D'Oro, Pierre-Luc Bacon

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

Comments Michel Ma and Pierluca D'Oro contributed equally

2401.00057 2024-01-02 cs.LG cs.CV 94%

Generalization properties of contrastive world models

Kandan Ramakrishnan, R. James Cotton, Xaq Pitkow, Andreas S. Tolias

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

Comments Accepted at the NeurIPS 2023 Workshop: Self-Supervised Learning - Theory and Practice

2312.17227 2023-12-29 cs.LG cs.AI 94%

Gradient-based Planning with World Models

Jyothir S, Siddhartha Jalagam, Yann LeCun, Vlad Sobal

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

2312.02019 2023-12-05 cs.LG cs.AI 94%

Action Inference by Maximising Evidence: Zero-Shot Imitation from Observation with World Models

Xingyuan Zhang, Philip Becker-Ehmck, Patrick van der Smagt, Maximilian Karl

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

Comments NeurIPS 2023

2310.18534 2023-12-05 cs.LG cs.AI 94%

Multi Time Scale World Models

Vaisakh Shaj, Saleh Gholam Zadeh, Ozan Demir, Luiz Ricardo Douat, Gerhard Neumann

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

Comments Accepted as spotlight at NeurIPS 2023

2305.10626 2023-10-31 cs.CL cs.AI cs.LG 94%

Language Models Meet World Models: Embodied Experiences Enhance Language Models

Jiannan Xiang, Tianhua Tao, Yi Gu, Tianmin Shu, Zirui Wang, Zichao Yang, Zhiting Hu

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

2308.07234 2023-08-15 cs.CV cs.RO 94%

UniWorld: Autonomous Driving Pre-training via World Models

Chen Min, Dawei Zhao, Liang Xiao, Yiming Nie, Bin Dai

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

Comments 8 pages, 5 figures. arXiv admin note: substantial text overlap with arXiv:2305.18829

2211.15944 2023-07-14 cs.LG cs.AI 94%

The Effectiveness of World Models for Continual Reinforcement Learning

Samuel Kessler, Mateusz Ostaszewski, Michał Bortkiewicz, Mateusz Żarski, Maciej Wołczyk, Jack Parker-Holder, Stephen J. Roberts, Piotr Miłoś

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

Comments Accepted at CoLLAs 2023, 21 pages, 15 figures

2301.04783 2023-05-24 cs.CV cs.RO 94%

Predictive World Models from Real-World Partial Observations

Robin Karlsson, Alexander Carballo, Keisuke Fujii, Kento Ohtani, Kazuya Takeda

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

Comments Best Paper Award at IEEE MOST 2023

2210.12278 2022-10-25 cs.RO cs.AI 94%

Sample Efficient Robot Learning with Structured World Models

Tuluhan Akbulut, Max Merlin, Shane Parr, Benedict Quartey, Skye Thompson

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

2208.04892 2022-08-10 cs.AI cs.LG 94%

Intrinsically Motivated Learning of Causal World Models

Louis Annabi

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

Comments 6 pages, 3 figures, IMOL 2022 workshop

2010.02193 2022-02-15 cs.LG cs.AI stat.ML 94%

Mastering Atari with Discrete World Models

Danijar Hafner, Timothy Lillicrap, Mohammad Norouzi, Jimmy Ba

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

Comments Published at ICLR 2021. Website: https://danijar.com/dreamerv2

2202.05333 2022-02-14 cs.RO cs.LG 94%

Factored World Models for Zero-Shot Generalization in Robotic Manipulation

Ondrej Biza, Thomas Kipf, David Klee, Robert Platt, Jan-Willem van de Meent, Lawson L. S. Wong

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

2112.07263 2021-12-15 cs.LG cs.AI 94%

Quantifying Multimodality in World Models

Andreas Sedlmeier, Michael Kölle, Robert Müller, Leo Baudrexel, Claudia Linnhoff-Popien

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

2106.09608 2021-10-22 cs.LG cs.AI cs.CL 94%

Learning Knowledge Graph-based World Models of Textual Environments

Prithviraj Ammanabrolu, Mark O. Riedl

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

Comments Camera read, in Proceedings of NeurIPS 2021 Main Conference

2010.05767 2021-03-03 cs.LG cs.AI stat.ML 94%

Smaller World Models for Reinforcement Learning

Jan Robine, Tobias Uelwer, Stefan Harmeling

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

2012.14228 2020-12-29 cs.LG cs.AI 94%

Causal World Models by Unsupervised Deconfounding of Physical Dynamics

Minne Li, Mengyue Yang, Furui Liu, Xu Chen, Zhitang Chen, Jun Wang

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

2010.15622 2020-10-30 stat.ML cs.AI cs.LG 94%

Low-Variance Policy Gradient Estimation with World Models

Michal Nauman, Floris Den Hengst

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

1911.12247 2020-01-07 stat.ML cs.AI cs.LG 94%

Contrastive Learning of Structured World Models

Thomas Kipf, Elise van der Pol, Max Welling

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

Comments ICLR 2020

2608.20065 2026-08-21 cs.LG 新提交 93%

Orthogonal JEPA: Factorized Predictive States for Latent World Models

正交JEPA：用于潜在世界模型的分解预测状态

Taoyong Cui, Pheng Ann Heng, Wanli Ouyang

机构 * The Chinese University of Hong Kong (CUHK)（香港中文大学（CUHK））

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

AI总结该研究提出正交JEPA框架，通过正交预测分解改进潜在世界模型，在多类任务实验中评估了其表示质量、预测、规划及长时程稳定性。

2608.15309 2026-08-18 cs.AI 新提交 93%

Physiological World Models for Human State Transitions

面向人体状态转换的生理世界模型

Chongyang Zhang, Rendong Wang, Hao Zheng, Hanwen Zhang, Yang Liu, Xiaolong Wei, Bin Chong

专题命中通用世界模型：world model(title,abstract)；world model(title,abstract)；world models(title)；world models(title)

AI总结本文提出生理世界模型（PWM），引入人体状态转换标记，构建含六项基准任务的框架，用于建模人体生理状态响应现实因素的变化，助力个性化健康管理等场景。

2608.12939 2026-08-14 cs.LG 新提交 93%

Diagnosing JEPA World Models with Action-Conditioned Predictive Consistency

用动作条件预测一致性诊断JEPA世界模型

Guo An, Zijing Wu, Honghua Dong, Yuhao Yan, Zixuan Gui, Haochong Chen, Shanzhao Ruan, Xiang Wang, Yurong Ling, Qi Tian

机构 * Huawei（华为）； University of Science and Technology of China（中国科学技术大学）； Zhejiang University（浙江大学）； Tsinghua University（清华大学）； Harbin Institute of Technology（哈尔滨工业大学）； Guangdong Laboratory of Artificial Intelligence and Digital Economy (SZ)（广东省人工智能与数字经济实验室（深圳））

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

AI总结本研究针对JEPAs世界模型易受视觉扰动影响的问题，提出动作条件预测一致性（ACPC）诊断方法，定义IR与SR指标，经四视觉控制任务实验验证其可预测扰动带来的预测及代价变化。

2608.09696 2026-08-14 cs.AI 版本更新 93%

Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models

模型发现智能体：用于数据高效发现机制世界模型的大语言模型辅助贝叶斯实验设计

Kevin Murphy

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

AI总结该研究提出模型发现智能体（MDA），结合LLM与贝叶斯机制，在少量干预下发现机制世界模型，在三类基准上实现数据高效模型学习与可靠干预预测的SOTA性能。

Comments v3: further improve app A (algorithm pseudocode), add new app G (further related work) - (main text unchanged)

2604.18064 2026-08-13 cs.AI 版本更新 93%

Towards Human Motion World Models via Executable Behaviour Representations

通过可执行模型理解人类行为

Rimvydas Rubavicius, Manisha Dubey, N. Siddharth, Subramanian Ramamoorthy

机构 * School of Informatics The University of Edinburgh（信息学院爱丁堡大学）

专题命中通用世界模型：world model(title,abstract)；world models(title,abstract)；world model(title,abstract)；world models(title,abstract)

AI总结本文提出EXACT语言，通过可执行神经符号模型分析人类动作，提升动作分割和异常检测的效率与直观性。

Comments Accepted in ECCV2026 Workshop

1. 通用世界模型 4308 篇

In-Context Symmetries: Self-Supervised Learning through Contextual World Models

MuDreamer: Learning Predictive World Models without Reconstruction

Safe Exploration Using Bayesian World Models and Log-Barrier Optimization

Zero-shot Safety Prediction for Autonomous Robots with Foundation World Models

World Models via Policy-Guided Trajectory Diffusion

WorldGPT: A Sora-Inspired Video AI Agent as Rich World Models from Text and Image Inputs

Do Transformer World Models Give Better Policy Gradients?

Generalization properties of contrastive world models

Gradient-based Planning with World Models

Action Inference by Maximising Evidence: Zero-Shot Imitation from Observation with World Models

Multi Time Scale World Models

Language Models Meet World Models: Embodied Experiences Enhance Language Models

UniWorld: Autonomous Driving Pre-training via World Models

The Effectiveness of World Models for Continual Reinforcement Learning

Predictive World Models from Real-World Partial Observations

Sample Efficient Robot Learning with Structured World Models

Intrinsically Motivated Learning of Causal World Models

Mastering Atari with Discrete World Models

Factored World Models for Zero-Shot Generalization in Robotic Manipulation

Quantifying Multimodality in World Models

Learning Knowledge Graph-based World Models of Textual Environments

Smaller World Models for Reinforcement Learning

Causal World Models by Unsupervised Deconfounding of Physical Dynamics

Low-Variance Policy Gradient Estimation with World Models

Contrastive Learning of Structured World Models

Orthogonal JEPA: Factorized Predictive States for Latent World Models

Physiological World Models for Human State Transitions

Diagnosing JEPA World Models with Action-Conditioned Predictive Consistency

Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models

Towards Human Motion World Models via Executable Behaviour Representations