Multi-agent Reinforcement Learning on SMAC MMM2 v1 (test)
81.4Win RateHIBCG
Evaluation Results
| Method | Links | |
|---|---|---|
| HIBCGUnit Composition=3-type, Training Budget=2M2026.05 | 81.4 | |
| CommFormerUnit Composition=3-type, Training Budget=2M2026.05 | 77.4 | |
| BVMEUnit Composition=3-type, Training Budget=2M2026.05 | 75.5 | |
| HIB-flatUnit Composition=3-type, Training Budget=2M2026.05 | 54.2 | |
| MAGIUnit Composition=3-type, Training Budget=2M2026.05 | 52.8 | |
| AIB-onlyUnit Composition=3-type, Training Budget=2M2026.05 | 50.2 | |
| GACGUnit Composition=3-type, Training Budget=2M2026.05 | 48.8 | |
| QMIXUnit Composition=3-type, Training Budget=2M2026.05 | 38.6 | |
| ExpoCommUnit Composition=3-type, Training Budget=2M2026.05 | 20.4 |