ResearchBenchmarksSafe Reinforcement Learning on mountain-car Static dynamicsFollow0Mean Shield Invocations per EpisodeDMPS-0.160.9223.08May 22, 2024Evaluation ResultsMethodMethodLinksMean Shield Invocations per EpisodeSD Shield Invocations per EpisodeMean Safety Violations per EpisodeSD Safety Violations per EpisodeDMPS2024.0500——MPS2024.0500——REVEL2024.0544——CPO2024.05——22.1PPO-lag2024.05——22.228.9TD32024.05——00