MrCoM: A Meta-Regularized World-Model Generalizing Across Multi-Scenarios
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866915606974955520 |
|---|---|
| author | Xiong, Xuantang Mu, Ni Xie, Runpeng Yang, Senhao Wang, Yaqing Wang, Lexiang Luan, Yao Li, Siyuan Xu, Shuang Yang, Yiqin Xu, Bo |
| author_facet | Xiong, Xuantang Mu, Ni Xie, Runpeng Yang, Senhao Wang, Yaqing Wang, Lexiang Luan, Yao Li, Siyuan Xu, Shuang Yang, Yiqin Xu, Bo |
| contents | Model-based reinforcement learning (MBRL) is a crucial approach to enhance the generalization capabilities and improve the sample efficiency of RL algorithms. However, current MBRL methods focus primarily on building world models for single tasks and rarely address generalization across different scenarios. Building on the insight that dynamics within the same simulation engine share inherent properties, we attempt to construct a unified world model capable of generalizing across different scenarios, named Meta-Regularized Contextual World-Model (MrCoM). This method first decomposes the latent state space into various components based on the dynamic characteristics, thereby enhancing the accuracy of world-model prediction. Further, MrCoM adopts meta-state regularization to extract unified representation of scenario-relevant information, and meta-value regularization to align world-model optimization with policy learning across diverse scenario objectives. We theoretically analyze the generalization error upper bound of MrCoM in multi-scenario settings. We systematically evaluate our algorithm's generalization ability across diverse scenarios, demonstrating significantly better performance than previous state-of-the-art methods. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_06252 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | MrCoM: A Meta-Regularized World-Model Generalizing Across Multi-Scenarios Xiong, Xuantang Mu, Ni Xie, Runpeng Yang, Senhao Wang, Yaqing Wang, Lexiang Luan, Yao Li, Siyuan Xu, Shuang Yang, Yiqin Xu, Bo Machine Learning Artificial Intelligence Model-based reinforcement learning (MBRL) is a crucial approach to enhance the generalization capabilities and improve the sample efficiency of RL algorithms. However, current MBRL methods focus primarily on building world models for single tasks and rarely address generalization across different scenarios. Building on the insight that dynamics within the same simulation engine share inherent properties, we attempt to construct a unified world model capable of generalizing across different scenarios, named Meta-Regularized Contextual World-Model (MrCoM). This method first decomposes the latent state space into various components based on the dynamic characteristics, thereby enhancing the accuracy of world-model prediction. Further, MrCoM adopts meta-state regularization to extract unified representation of scenario-relevant information, and meta-value regularization to align world-model optimization with policy learning across diverse scenario objectives. We theoretically analyze the generalization error upper bound of MrCoM in multi-scenario settings. We systematically evaluate our algorithm's generalization ability across diverse scenarios, demonstrating significantly better performance than previous state-of-the-art methods. |
| title | MrCoM: A Meta-Regularized World-Model Generalizing Across Multi-Scenarios |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2511.06252 |