ResearchBenchmarksSafe Reinforcement Learning on road2d Static dynamicsSafe Reinforcement Learning on road2d Static dynamicsFollow0Shield Invocations / Episode (Mean)DMPS-0.001-0.000500.0005May 22, 2024Evaluation resultsMethodMethodLinksShield Invocations / Episode (Mean)Shield Invocations / Episode (SD)Safety Violations / Episode (Mean)Safety Violations / Episode (SD)DMPS2024.0500——MPS2024.0500——REVEL2024.0500——CPO2024.05——00PPO-lag2024.05——00TD32024.05——00