Adversarial Attack on Breakout White-box discrete (test)
65.73Cumulative RewardPA-AD
Evaluation Results
| Method | Links | |
|---|---|---|
| PA-ADTarget Policy=PPO2026.02 | 65.73 | |
| S-TTarget Policy=TRPO2026.02 | 58.47 | |
| S-TTarget Policy=DQN2026.02 | 57.68 | |
| S-TTarget Policy=A3C2026.02 | 55.44 | |
| PA-ADTarget Policy=A3C2026.02 | 53.87 | |
| PIATarget Policy=PPO2026.02 | 53.47 | |
| EDGETarget Policy=PPO2026.02 | 50.33 | |
| PA-ADTarget Policy=TRPO2026.02 | 47.96 | |
| PA-ADTarget Policy=DQN2026.02 | 46.35 | |
| EDGETarget Policy=A3C2026.02 | 46.32 | |
| EDGETarget Policy=TRPO2026.02 | 44.89 | |
| S-TTarget Policy=PPO2026.02 | 44.36 | |
| EDGETarget Policy=DQN2026.02 | 43.26 | |
| FGSMTarget Policy=PPO2026.02 | 40.56 | |
| PIATarget Policy=TRPO2026.02 | 40.51 | |
| SkipTarget Policy=TRPO2026.02 | 38.79 | |
| SkipTarget Policy=PPO2026.02 | 38.42 | |
| AdvRL-GANTarget Policy=PPO2026.02 | 38.21 | |
| FGSMTarget Policy=A3C2026.02 | 35.45 | |
| PIATarget Policy=DQN2026.02 | 35.17 | |
| FGSMTarget Policy=TRPO2026.02 | 34.58 | |
| TSGETarget Policy=PPO2026.02 | 34.32 | |
| FGSMTarget Policy=DQN2026.02 | 34.3 | |
| AdvRL-GANTarget Policy=TRPO2026.02 | 33.98 | |
| SkipTarget Policy=A3C2026.02 | 32.46 | |
| AdvRL-GANTarget Policy=DQN2026.02 | 32.44 | |
| SkipTarget Policy=DQN2026.02 | 31.45 | |
| PIATarget Policy=A3C2026.02 | 27.84 | |
| TSGETarget Policy=TRPO2026.02 | 25.46 | |
| TSGETarget Policy=A3C2026.02 | 15.29 | |
| AdvRL-GANTarget Policy=A3C2026.02 | 9.11 | |
| AATTarget Policy=A3C2026.02 | 8.21 | |
| AATTarget Policy=PPO2026.02 | 6.88 | |
| AATTarget Policy=DQN2026.02 | 6.56 | |
| TSGETarget Policy=DQN2026.02 | 6.42 | |
| AATTarget Policy=TRPO2026.02 | 5.98 |