Multi-Agent Reinforcement Learning (Predator-Prey) on MPE PP_6/2
685.5Average Cumulative RewardWALL
Evaluation results
| Method | Links | |
|---|---|---|
| WALLAttack Method=EGA2025.02 | 685.5 | |
| WALLAttack Method=Natural2025.02 | 675 | |
| RANDOMAttack Method=Natural2025.02 | 663.8 | |
| WALLAttack Method=Radom Attack2025.02 | 654 | |
| ROMANCEAttack Method=Natural2025.02 | 648.6 | |
| RANDOMAttack Method=Radom Attack2025.02 | 638.7 | |
| ROMANCEAttack Method=Radom Attack2025.02 | 624.2 | |
| ROMANCEAttack Method=EGA2025.02 | 605.1 | |
| RANDOMAttack Method=EGA2025.02 | 604.6 | |
| WALLAttack Method=Wolfpack Adversarial Attack2025.02 | 599 | |
| RANDOMAttack Method=Wolfpack Adversarial Attack2025.02 | 571 | |
| ROMANCEAttack Method=Wolfpack Adversarial Attack2025.02 | 554.6 | |
| Vanilla QMIXAttack Method=Natural2025.02 | 538.8 | |
| Vanilla QMIXAttack Method=Radom Attack2025.02 | 522.6 | |
| Vanilla QMIXAttack Method=EGA2025.02 | 499.8 | |
| Vanilla QMIXAttack Method=Wolfpack Adversarial Attack2025.02 | 415.6 |