Feasibility Consistent Representation Learning for Safe Reinforcement Learning (ICML 2024). Current SOTA model-free safe RL algorithm on safety-gym
Primary LanguagePythonApache License 2.0Apache-2.0