ResearchBenchmarksSafe Reinforcement Learning on Safety Gymnasium AntFollow4.591Time To Safety (TTS)APPO3.815129.0523114.289519.52669Jun 12, 2026Evaluation ResultsMethodMethodLinksTime To Safety (TTS)Risk Profile (RP)Number of Violations (#V)APPO2026.064.5910.9975138.6CSPO2026.065.1511.001122.6CPPOPID2026.068.871.0046134.8PPO-Lag2026.0614.0450.9979122CUP2026.0621.3881.0229124.8FOCOPS2026.0623.9881.033150.4