Agentic World Modeling: Foundations, Capabilities, Laws, and Beyond
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915955542589440 |
|---|---|
| author | Chu, Meng Zhang, Xuan Billy Lin, Kevin Qinghong Kong, Lingdong Zhang, Jize Tu, Teng Ma, Weijian Huang, Ziqi Yang, Senqiao Huang, Wei Jin, Yeying Rao, Zhefan Ye, Jinhui Lin, Xinyu Zhang, Xichen Hu, Qisheng Yang, Shuai Shen, Leyang Chow, Wei Dong, Yifei Wu, Fengyi Long, Quanyu Xia, Bin Yu, Shaozuo Zhu, Mingkang Zhang, Wenhu Huang, Jiehui Gui, Haokun Che, Haoxuan Chen, Long Chen, Qifeng Zhang, Wenxuan Wang, Wenya Qi, Xiaojuan Deng, Yang Li, Yanwei Shou, Mike Zheng Cheng, Zhi-Qi Ng, See-Kiong Liu, Ziwei Torr, Philip Jia, Jiaya |
| author_facet | Chu, Meng Zhang, Xuan Billy Lin, Kevin Qinghong Kong, Lingdong Zhang, Jize Tu, Teng Ma, Weijian Huang, Ziqi Yang, Senqiao Huang, Wei Jin, Yeying Rao, Zhefan Ye, Jinhui Lin, Xinyu Zhang, Xichen Hu, Qisheng Yang, Shuai Shen, Leyang Chow, Wei Dong, Yifei Wu, Fengyi Long, Quanyu Xia, Bin Yu, Shaozuo Zhu, Mingkang Zhang, Wenhu Huang, Jiehui Gui, Haokun Che, Haoxuan Chen, Long Chen, Qifeng Zhang, Wenxuan Wang, Wenya Qi, Xiaojuan Deng, Yang Li, Yanwei Shou, Mike Zheng Cheng, Zhi-Qi Ng, See-Kiong Liu, Ziwei Torr, Philip Jia, Jiaya |
| contents | As AI systems move from generating text to accomplishing goals through sustained interaction, the ability to model environment dynamics becomes a central bottleneck. Agents that manipulate objects, navigate software, coordinate with others, or design experiments require predictive environment models, yet the term world model carries different meanings across research communities. We introduce a "levels x laws" taxonomy organized along two axes. The first defines three capability levels: L1 Predictor, which learns one-step local transition operators; L2 Simulator, which composes them into multi-step, action-conditioned rollouts that respect domain laws; and L3 Evolver, which autonomously revises its own model when predictions fail against new evidence. The second identifies four governing-law regimes: physical, digital, social, and scientific. These regimes determine what constraints a world model must satisfy and where it is most likely to fail. Using this framework, we synthesize over 400 works and summarize more than 100 representative systems spanning model-based reinforcement learning, video generation, web and GUI agents, multi-agent social simulation, and AI-driven scientific discovery. We analyze methods, failure modes, and evaluation practices across level-regime pairs, propose decision-centric evaluation principles and a minimal reproducible evaluation package, and outline architectural guidance, open problems, and governance challenges. The resulting roadmap connects previously isolated communities and charts a path from passive next-step prediction toward world models that can simulate, and ultimately reshape, the environments in which agents operate. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_22748 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Agentic World Modeling: Foundations, Capabilities, Laws, and Beyond Chu, Meng Zhang, Xuan Billy Lin, Kevin Qinghong Kong, Lingdong Zhang, Jize Tu, Teng Ma, Weijian Huang, Ziqi Yang, Senqiao Huang, Wei Jin, Yeying Rao, Zhefan Ye, Jinhui Lin, Xinyu Zhang, Xichen Hu, Qisheng Yang, Shuai Shen, Leyang Chow, Wei Dong, Yifei Wu, Fengyi Long, Quanyu Xia, Bin Yu, Shaozuo Zhu, Mingkang Zhang, Wenhu Huang, Jiehui Gui, Haokun Che, Haoxuan Chen, Long Chen, Qifeng Zhang, Wenxuan Wang, Wenya Qi, Xiaojuan Deng, Yang Li, Yanwei Shou, Mike Zheng Cheng, Zhi-Qi Ng, See-Kiong Liu, Ziwei Torr, Philip Jia, Jiaya Artificial Intelligence As AI systems move from generating text to accomplishing goals through sustained interaction, the ability to model environment dynamics becomes a central bottleneck. Agents that manipulate objects, navigate software, coordinate with others, or design experiments require predictive environment models, yet the term world model carries different meanings across research communities. We introduce a "levels x laws" taxonomy organized along two axes. The first defines three capability levels: L1 Predictor, which learns one-step local transition operators; L2 Simulator, which composes them into multi-step, action-conditioned rollouts that respect domain laws; and L3 Evolver, which autonomously revises its own model when predictions fail against new evidence. The second identifies four governing-law regimes: physical, digital, social, and scientific. These regimes determine what constraints a world model must satisfy and where it is most likely to fail. Using this framework, we synthesize over 400 works and summarize more than 100 representative systems spanning model-based reinforcement learning, video generation, web and GUI agents, multi-agent social simulation, and AI-driven scientific discovery. We analyze methods, failure modes, and evaluation practices across level-regime pairs, propose decision-centric evaluation principles and a minimal reproducible evaluation package, and outline architectural guidance, open problems, and governance challenges. The resulting roadmap connects previously isolated communities and charts a path from passive next-step prediction toward world models that can simulate, and ultimately reshape, the environments in which agents operate. |
| title | Agentic World Modeling: Foundations, Capabilities, Laws, and Beyond |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2604.22748 |