Communication-based Multi-agent Reinforcement Learning on SH Clean
37.2Average Winning RateSHINE
Evaluation Results
| Method | Links | |
|---|---|---|
| SHINE2026.05 | 37.2 | |
| BehaviorGuard2026.05 | 33.3 | |
| Original2026.05 | 33 | |
| BIRD2026.05 | 31.3 | |
| Direct retraining2026.05 | 27.2 |
Loading the SOTA2 catalog…