Multi-Agent Reinforcement Learning on SMAC 2s3z StarCraft II (test)
99.5Natural AccIBAL
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| IBAL2026.05 | 99.5 | 93.3 | 89.8 | 89.8 | 83.4 | 78.7 | 89.1 | |
| WALL2026.05 | 98.9 | 91.7 | 86.7 | 87.5 | 92.7 | 13.5 | 78.9 | |
| Rand-Obs2026.05 | 98.2 | 80 | 89.3 | 53.1 | 52.1 | 5.2 | 63 | |
| ROMANCE2026.05 | 97.9 | 82 | 76.7 | 83.3 | 65 | 13.3 | 69.7 | |
| Vanilla QMIX2026.05 | 97.6 | 75 | 58.3 | 53.8 | 39.2 | 3.1 | 54.7 | |
| ERNIE2026.05 | 97.6 | 82.3 | 74.7 | 63.5 | 49.2 | 4.2 | 61.9 | |
| ATLA2026.05 | 96.6 | 80.3 | 89.7 | 54.2 | 53.1 | 9.4 | 63.9 | |
| FGSM2026.05 | 96.5 | 85.7 | 95.7 | 62.5 | 61.4 | 6.2 | 68 | |
| Rand-Act2026.05 | 96.1 | 79 | 77.3 | 68.7 | 62.6 | 4.1 | 64.7 |