faucet-open on Meta-World v2 (test)
4,383Best Attack RewardATLA-PPO
Evaluation Results
| Method | Links | |
|---|---|---|
| ATLA-PPODefense mechanism=AdvTraining2024.06 | 4,383 | |
| RAD-PPODefense mechanism=Regret2024.06 | 4,254 | |
| PPODefense mechanism=No defense2024.06 | 3,031 | |
| WocaR-PPODefense mechanism=Partial smoothing2024.06 | 3,012 | |
| PA-ATLA-PPODefense mechanism=AdvTraining2024.06 | 2,358 | |
| TDRT-PPODefense mechanism=Smoothing, Time-discounting=true2024.06 | 1,942 | |
| SA-PPODefense mechanism=Smoothing, Time-discounting=false2024.06 | 1,763 |