door-lock on Meta-World v2 (test)
2,043Best Attack RewardPPO
Evaluation results
| Method | Links | |
|---|---|---|
| PPODefense mechanism=No defense2024.06 | 2,043 | |
| ATLA-PPODefense mechanism=AdvTraining2024.06 | 1,020 | |
| PA-ATLA-PPODefense mechanism=AdvTraining2024.06 | 992 | |
| RAD-PPODefense mechanism=Regret2024.06 | 712 | |
| WocaR-PPODefense mechanism=Partial smoothing2024.06 | 562 | |
| TDRT-PPODefense mechanism=Smoothing, Time-discounting=true2024.06 | 487 | |
| SA-PPODefense mechanism=Smoothing, Time-discounting=false2024.06 | 478 |