ResearchBenchmarksSafe Reinforcement Learning on SafetyGym PointGoal1 (PG1)Follow19.6Final RewardITES11.69613.74815.817.852Jun 21, 2026Evaluation ResultsMethodMethodLinksFinal RewardFinal CostITES2026.0619.632.83TD3LAG2026.0615.7556.22FOCOPS2026.0614.9733.72CUP2026.0614.4219.02PPOLAG2026.0612.9625.8SafeDreamer2026.061210.3