Multi-Agent Reinforcement Learning (Predator-Prey) on MPE PP_9/3
802.5Average Cumulative RewardWALL
Evaluation results
| Method | Links | |
|---|---|---|
| WALLAttack Method=Natural2025.02 | 802.5 | |
| WALLAttack Method=Radom Attack2025.02 | 802.5 | |
| ROMANCEAttack Method=Natural2025.02 | 721.4 | |
| ROMANCEAttack Method=Radom Attack2025.02 | 701.7 | |
| WALLAttack Method=Wolfpack Adversarial Attack2025.02 | 698.1 | |
| WALLAttack Method=EGA2025.02 | 682.7 | |
| RANDOMAttack Method=Natural2025.02 | 666.9 | |
| Vanilla QMIXAttack Method=Natural2025.02 | 661.9 | |
| RANDOMAttack Method=Radom Attack2025.02 | 657.9 | |
| Vanilla QMIXAttack Method=Radom Attack2025.02 | 656.4 | |
| RANDOMAttack Method=Wolfpack Adversarial Attack2025.02 | 624.3 | |
| ROMANCEAttack Method=EGA2025.02 | 594.4 | |
| RANDOMAttack Method=EGA2025.02 | 589.7 | |
| Vanilla QMIXAttack Method=EGA2025.02 | 585.1 | |
| ROMANCEAttack Method=Wolfpack Adversarial Attack2025.02 | 570.3 | |
| Vanilla QMIXAttack Method=Wolfpack Adversarial Attack2025.02 | 551.7 |