ResearchBenchmarksSafe Reinforcement Learning on SafetyGym PointGoal1 sparseFollow8.64Final RewardITES-0.27282.04114.3556.6689Jun 21, 2026Evaluation ResultsMethodMethodLinksFinal RewardFinal CostITES2026.068.6433.46FOCOPS2026.067.63522.9PPOLAG2026.066.8532.645MBPPOL2026.066.1534.1CUP2026.061.75512.925SafeDreamer2026.060.53.22TD3LAG2026.060.0764.7