Offline Reinforcement Learning on Ising Model N=1000 (Medium)
87.9Normalized ReturnMF-Diffuser
Evaluation results
| Method | Links | |
|---|---|---|
| MF-DiffuserN=1000, Seeds=5, Online rollouts=102026.05 | 87.9 | |
| MF-CDMs-RLN=1000, Seeds=5, Online rollouts=102026.05 | 84.6 | |
| MFQ-OfflineN=1000, Seeds=5, Online rollouts=102026.05 | 83.5 | |
| OryxN=1000, Seeds=5, Online rollouts=102026.05 | 82 | |
| DoFN=1000, Seeds=5, Online rollouts=102026.05 | 78.8 | |
| Indep. DiffuserN=1000, Seeds=5, Online rollouts=102026.05 | 74.3 | |
| MADiffN=1000, Seeds=5, Online rollouts=102026.05 | 64.2 | |
| Joint DiffuserN=1000, Seeds=5, Online rollouts=102026.05 | 58.7 | |
| OMARN=1000, Seeds=5, Online rollouts=10, mean-field adaptation=true2026.05 | 53.1 | |
| MA-TD3+BCN=1000, Seeds=5, Online rollouts=10, mean-field adaptation=true2026.05 | 51.8 |