Safety-constrained Reinforcement Learning on Safety-Gym SafetyPointGoal1 (evaluation)
11.37Average RewardPPO
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| PPO2026.05 | 11.37 | 109.87 | 1 | 4.395 | 0.103 | |
| PPO-Lag2026.05 | 8.14 | 90.27 | 0.866 | 3.611 | 0.09 | |
| RiskGated2026.05 | 5.39 | 36.93 | 0.474 | 1.477 | 0.146 | |
| TRPO-Lag2026.05 | 2.99 | 195.33 | 0.318 | 7.813 | 0.015 | |
| RCPO2026.05 | 2.99 | 195.33 | 0.318 | 7.813 | 0.015 | |
| CUP2026.05 | 2.3 | 41.67 | 0.245 | 1.667 | 0.055 | |
| FOCOPS2026.05 | 2.26 | 309.33 | 0.241 | 12.373 | 0.007 | |
| CPPO-PID2026.05 | 0.68 | 110.67 | 0.072 | 4.427 | 0.006 | |
| CPO2026.05 | -0.12 | 25.67 | -0.013 | 1.027 | -0.005 | |
| POMDP2026.05 | -0.71 | 101.5 | -0.063 | 4.06 | -0.007 | |
| PCPO2026.05 | -0.73 | 5.67 | -0.077 | 0.227 | -0.129 |