Multi-agent Reinforcement Learning on SMAC 1c3s5z v1 (test)
91.4Test Win RateQMIX
Evaluation Results
| Method | Links | |
|---|---|---|
| QMIXUnit Composition=3-type, Training Budget=2M2026.05 | 91.4 | |
| AIB-onlyUnit Composition=3-type, Training Budget=2M2026.05 | 91.4 | |
| BVMEUnit Composition=3-type, Training Budget=2M2026.05 | 91.3 | |
| HIBCGUnit Composition=3-type, Training Budget=2M2026.05 | 91.2 | |
| HIB-flatUnit Composition=3-type, Training Budget=2M2026.05 | 90.6 | |
| MAGIUnit Composition=3-type, Training Budget=2M2026.05 | 90.5 | |
| GACGUnit Composition=3-type, Training Budget=2M2026.05 | 89.7 | |
| ExpoCommUnit Composition=3-type, Training Budget=2M2026.05 | 89.5 | |
| CommFormerUnit Composition=3-type, Training Budget=2M2026.05 | 87.8 |