Window Open on Meta-World v2 (test)
671Best Attack RewardPA-ATLA-PPO
Evaluation Results
| Method | Links | |
|---|---|---|
| PA-ATLA-PPODefense mechanism=AdvTraining2024.06 | 671 | |
| ATLA-PPODefense mechanism=AdvTraining2024.06 | 586 | |
| PPODefense mechanism=No defense2024.06 | 566 | |
| RAD-PPODefense mechanism=Regret2024.06 | 501 | |
| WocaR-PPODefense mechanism=Partial smoothing2024.06 | 295 | |
| SA-PPODefense mechanism=Smoothing, Time-discounting=false2024.06 | 272 | |
| TDRT-PPODefense mechanism=Smoothing, Time-discounting=true2024.06 | 254 |