Multi-Agent Reinforcement Learning on SMAC 1c3s5z v1
99.1Natural PerformanceROMANCE
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| ROMANCE2026.05 | 99.1 | 94.3 | 91 | 89.6 | 89.2 | 12.5 | 79.2 | |
| IBAL2026.05 | 99 | 97.3 | 98.3 | 93.9 | 98.9 | 91.3 | 96.4 | |
| WALL2026.05 | 98.9 | 96.3 | 96.7 | 93.8 | 95.8 | 18.8 | 83.4 | |
| Vanilla QMIX2026.05 | 98.8 | 84.7 | 81.7 | 83.8 | 52.7 | 18.8 | 70.7 | |
| FGSM2026.05 | 98.6 | 92.7 | 97 | 88.5 | 67.6 | 6.3 | 75.8 | |
| Rand-Act2026.05 | 98.3 | 90 | 83.3 | 86.5 | 64.8 | 6.3 | 71.5 | |
| ATLA2026.05 | 96.9 | 94 | 94.7 | 84.4 | 73.1 | 24 | 77.1 | |
| Rand-Obs2026.05 | 96.8 | 89 | 86.3 | 86.8 | 56.1 | 7.3 | 70.9 | |
| ERNIE2026.05 | 94.8 | 93.3 | 93 | 88.5 | 80.6 | 17.7 | 77.2 |