Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2606.00133 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910273006206976 |
|---|---|
| author | Zidan, Arif Hassan Pan, Yi Jiang, Hanqi Yan, Ruiyu Ruan, Wei Wu, Zihao Chen, Lifeng You, Weihang Li, Xinliang Chen, Bowen Hu, Huawen Wang, Peilong Liu, Sizhuang Zhang, Jing Li, Siyuan Liu, Zhengliang Bao, Yu Zhao, Lin Sun, Lichao Zhu, Dajiang Li, Xiang Lv, Jinglei Li, Quanzheng Liu, Wei Liu, Tianming Zhang, Wei |
| author_facet | Zidan, Arif Hassan Pan, Yi Jiang, Hanqi Yan, Ruiyu Ruan, Wei Wu, Zihao Chen, Lifeng You, Weihang Li, Xinliang Chen, Bowen Hu, Huawen Wang, Peilong Liu, Sizhuang Zhang, Jing Li, Siyuan Liu, Zhengliang Bao, Yu Zhao, Lin Sun, Lichao Zhu, Dajiang Li, Xiang Lv, Jinglei Li, Quanzheng Liu, Wei Liu, Tianming Zhang, Wei |
| contents | World models, internal simulators that learn the structure and dynamics of an environment, have emerged as a central paradigm in the pursuit of artificial general intelligence, enabling agents to predict, plan, and reason within learned representations. Despite rapid progress across reinforcement learning, robotics, autonomous driving, and video generation, the field lacks a unified framework integrating its diverse architectural choices, training methods, reasoning mechanisms, and application settings. This survey addresses that gap with a multi-axis taxonomy organized along four dimensions: (i) architecture, encompassing representation format, dynamics formulation, input modality, learning paradigm, and downstream application; (ii) methodological family, including state-space and recurrent approaches, transformer-based models, diffusion-based generators, physics-informed networks, and language-augmented multimodal systems; (iii) reasoning strategy, covering imagination-based planning, latent policy learning, counterfactual reasoning, and planning under uncertainty; and (iv) application domain, spanning robotics, autonomous driving, video prediction, multimodal agents, reinforcement learning, scientific modeling, medical imaging, educational measurement, and business and finance. Tracing the field from early cognitive-science foundations to milestone systems such as PlaNet, the Dreamer family, MuZero, Sora, Cosmos, and Genie, we examine how these dimensions interact and highlight the recent convergence of chain-of-thought reasoning with world-model imagination. We review evaluation protocols and benchmarks, identify persistent challenges such as compounding prediction errors, sim-to-real transfer, and fragmented evaluation, and outline future directions toward unified multimodal world models, foundation-scale interactive simulators, and safe deployment in safety-critical domains. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2606_00133 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | World Models: A Comprehensive Survey of Architectures, Methodologies, Reasoning Paradigms, and Applications Zidan, Arif Hassan Pan, Yi Jiang, Hanqi Yan, Ruiyu Ruan, Wei Wu, Zihao Chen, Lifeng You, Weihang Li, Xinliang Chen, Bowen Hu, Huawen Wang, Peilong Liu, Sizhuang Zhang, Jing Li, Siyuan Liu, Zhengliang Bao, Yu Zhao, Lin Sun, Lichao Zhu, Dajiang Li, Xiang Lv, Jinglei Li, Quanzheng Liu, Wei Liu, Tianming Zhang, Wei Machine Learning Emerging Technologies World models, internal simulators that learn the structure and dynamics of an environment, have emerged as a central paradigm in the pursuit of artificial general intelligence, enabling agents to predict, plan, and reason within learned representations. Despite rapid progress across reinforcement learning, robotics, autonomous driving, and video generation, the field lacks a unified framework integrating its diverse architectural choices, training methods, reasoning mechanisms, and application settings. This survey addresses that gap with a multi-axis taxonomy organized along four dimensions: (i) architecture, encompassing representation format, dynamics formulation, input modality, learning paradigm, and downstream application; (ii) methodological family, including state-space and recurrent approaches, transformer-based models, diffusion-based generators, physics-informed networks, and language-augmented multimodal systems; (iii) reasoning strategy, covering imagination-based planning, latent policy learning, counterfactual reasoning, and planning under uncertainty; and (iv) application domain, spanning robotics, autonomous driving, video prediction, multimodal agents, reinforcement learning, scientific modeling, medical imaging, educational measurement, and business and finance. Tracing the field from early cognitive-science foundations to milestone systems such as PlaNet, the Dreamer family, MuZero, Sora, Cosmos, and Genie, we examine how these dimensions interact and highlight the recent convergence of chain-of-thought reasoning with world-model imagination. We review evaluation protocols and benchmarks, identify persistent challenges such as compounding prediction errors, sim-to-real transfer, and fragmented evaluation, and outline future directions toward unified multimodal world models, foundation-scale interactive simulators, and safe deployment in safety-critical domains. |
| title | World Models: A Comprehensive Survey of Architectures, Methodologies, Reasoning Paradigms, and Applications |
| topic | Machine Learning Emerging Technologies |
| url | https://arxiv.org/abs/2606.00133 |