Continuous Control on Safety-Gymnasium SafetyPointGoal1 (generalization)
24.5Episodic RewardCPO-FOAM
Evaluation results
| Method | Links | ||
|---|---|---|---|
| CPO-FOAM2026.04 | 24.5 | 8.5 | |
| Lagrangian PPO2026.04 | 18.5 | 38.4 | |
| Vanilla CPO2026.04 | 17.4 | 24.5 | |
| RCPO2026.04 | 17.2 | 32.5 | |
| PID-Lagrangian2026.04 | 16.8 | 18.5 | |
| IPO2026.04 | 16.5 | 28.6 |