Safe Reinforcement Learning on Bullet Safety Gym AntRun
684.8RewardFOCOPS
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| FOCOPSKnowledge Level=Oracle2025.04 | 684.8 | 23.3 | — | |
| PPO-LagrangianKnowledge Level=Oracle2025.04 | 683.7 | 25.4 | — | |
| CPPO-PIDKnowledge Level=Oracle2025.04 | 675.5 | 22.6 | — | |
| C-T BaselineKnowledge Level=Zero Knowledge2025.04 | 654.6 | 33.3 | 6,000 | |
| TRPO-SauteKnowledge Level=Oracle2025.04 | 652 | 13.8 | — | |
| CUPKnowledge Level=Oracle2025.04 | 649.1 | 55.2 | — | |
| TraCeSKnowledge Level=Zero Knowledge2025.04 | 590.5 | 21.4 | 6,521 | |
| TraCeSKnowledge Level=Zero Knowledge2025.04 | 590.5 | 21.4 | 6,521 |