SafeDreamer: Safe Reinforcement Learning with World Models
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2023
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866911981034799104 |
|---|---|
| author | Huang, Weidong Ji, Jiaming Xia, Chunhe Zhang, Borong Yang, Yaodong |
| author_facet | Huang, Weidong Ji, Jiaming Xia, Chunhe Zhang, Borong Yang, Yaodong |
| contents | The deployment of Reinforcement Learning (RL) in real-world applications is constrained by its failure to satisfy safety criteria. Existing Safe Reinforcement Learning (SafeRL) methods, which rely on cost functions to enforce safety, often fail to achieve zero-cost performance in complex scenarios, especially vision-only tasks. These limitations are primarily due to model inaccuracies and inadequate sample efficiency. The integration of the world model has proven effective in mitigating these shortcomings. In this work, we introduce SafeDreamer, a novel algorithm incorporating Lagrangian-based methods into world model planning processes within the superior Dreamer framework. Our method achieves nearly zero-cost performance on various tasks, spanning low-dimensional and vision-only input, within the Safety-Gymnasium benchmark, showcasing its efficacy in balancing performance and safety in RL tasks. Further details can be found in the code repository: \url{https://github.com/PKU-Alignment/SafeDreamer}. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2307_07176 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | SafeDreamer: Safe Reinforcement Learning with World Models Huang, Weidong Ji, Jiaming Xia, Chunhe Zhang, Borong Yang, Yaodong Machine Learning Artificial Intelligence Computer Vision and Pattern Recognition The deployment of Reinforcement Learning (RL) in real-world applications is constrained by its failure to satisfy safety criteria. Existing Safe Reinforcement Learning (SafeRL) methods, which rely on cost functions to enforce safety, often fail to achieve zero-cost performance in complex scenarios, especially vision-only tasks. These limitations are primarily due to model inaccuracies and inadequate sample efficiency. The integration of the world model has proven effective in mitigating these shortcomings. In this work, we introduce SafeDreamer, a novel algorithm incorporating Lagrangian-based methods into world model planning processes within the superior Dreamer framework. Our method achieves nearly zero-cost performance on various tasks, spanning low-dimensional and vision-only input, within the Safety-Gymnasium benchmark, showcasing its efficacy in balancing performance and safety in RL tasks. Further details can be found in the code repository: \url{https://github.com/PKU-Alignment/SafeDreamer}. |
| title | SafeDreamer: Safe Reinforcement Learning with World Models |
| topic | Machine Learning Artificial Intelligence Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2307.07176 |