Communication-based Multi-agent Reinforcement Learning on YSNP Clean
51Average Win RateBehaviorGuard
Evaluation results
| Method | Links | |
|---|---|---|
| BehaviorGuard2026.05 | 51 | |
| Original2026.05 | 48.7 | |
| SHINE2026.05 | 48.4 | |
| BIRD2026.05 | 44.1 | |
| Direct retraining2026.05 | 37.3 |
| Method | Links | |
|---|---|---|
| BehaviorGuard2026.05 | 51 | |
| Original2026.05 | 48.7 | |
| SHINE2026.05 | 48.4 | |
| BIRD2026.05 | 44.1 | |
| Direct retraining2026.05 | 37.3 |