Communication-based Multi-agent Reinforcement Learning (com-MARL) on RTGA Poisoned
49.7Average Winning RateBehaviorGuard
Evaluation results
| Method | Links | |
|---|---|---|
| BehaviorGuard2026.05 | 49.7 | |
| SHINE2026.05 | 45.2 | |
| BIRD2026.05 | 42.4 | |
| Direct retraining2026.05 | 37.4 | |
| Original2026.05 | 18.8 |