SafeDreamer: Safe Reinforcement Learning with World Models

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: https://github.com/PKU-Alignment/SafeDreamer . INTRODUCTION A challenge in the real-world deployment of RL agents is to prevent unsafe situations (Feng et al., 2023; Ji et al., 2023a). SafeRL proposes a practical solution by defining a constrained Markov decision process (CMDP) (Altman, 1999) and integrating an additional cost function to quantify potential hazardous behaviors. In this process, agents aim to maximize rewards while maintaining costs below predefined constraint thresholds. Several remarkable algorithms have been developed on this foundation (