Safe Reinforcement Learning on RacecarGoal
34.03Violation RateTRPO
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| TRPOSafety Mechanism=ADRC2026.01 | 34.03 | 6.16 | 22.02 | |
| CPPOSafety Mechanism=ADRC2026.01 | 47.08 | 12.31 | 30.12 | |
| TRPOSafety Mechanism=PID2026.01 | 49.33 | 11.7 | 30.94 | |
| TRPOSafety Mechanism=Lagrangian2026.01 | 64.31 | 22.89 | 43.67 | |
| DDPGSafety Mechanism=ADRC2026.01 | 68.41 | 17.81 | 39.41 | |
| TD3Safety Mechanism=ADRC2026.01 | 71.24 | 17.55 | 39.71 | |
| DDPGSafety Mechanism=Lagrangian2026.01 | 71.44 | 18.47 | 40.76 | |
| DDPGSafety Mechanism=PID2026.01 | 72.05 | 17.93 | 40.29 | |
| CPPOSafety Mechanism=PID2026.01 | 72.3 | 27.11 | 49.02 | |
| TD3Safety Mechanism=PID2026.01 | 73.31 | 19.19 | 41.57 | |
| TD3Safety Mechanism=Lagrangian2026.01 | 75.26 | 19.93 | 42.57 | |
| CPPOSafety Mechanism=Lagrangian2026.01 | 80.87 | 31.18 | 54.24 |