Multi-Agent Reinforcement Learning on MPE Speaker-Listener
27.9ReturnVDN
Evaluation results
| Method | Links | ||
|---|---|---|---|
| VDN2026.04 | 27.9 | — | |
| IQL2026.04 | 27.8 | — | |
| DICG2026.04 | 23.2 | — | |
| DCG2026.04 | 21.2 | — | |
| QMIX2026.04 | 18.6 | — | |
| LLM-Priors2026.04 | 16.9 | — | |
| MAPPOObservation Setting=FO2026.04 | -46 | 6 | |
| KD-MARLObservation Setting=LH2026.04 | -48 | 4 | |
| KD-MARLObservation Setting=LH+A2026.04 | -50 | 3.9 | |
| QMIXObservation Setting=FO2026.04 | -55 | 4.8 | |
| MAPPOObservation Setting=LH2026.04 | -82 | 5.3 | |
| VDNObservation Setting=FO2026.04 | -90 | 4.7 | |
| MAPPOObservation Setting=LH+A2026.04 | -118 | 4 | |
| QMIXObservation Setting=LH2026.04 | -138 | 4.6 | |
| VDNObservation Setting=LH2026.04 | -170 | 4.5 | |
| QMIXObservation Setting=LH+A2026.04 | -205 | 4.2 | |
| VDNObservation Setting=LH+A2026.04 | -240 | 4.3 |