Finding epsilon-approximate equilibrium on Robust Markov Games Generative Model setting
3Sample ComplexityDR-NVI
Evaluation results
| Method | Links | |
|---|---|---|
| DR-NVIRobust MG Model=Tabular, Uncertainty Set=(s, a)-rectangularity, Break Curse=No2026.05 | 3 | |
| P2MPORobust MG Model=Tabular, Uncertainty Set=(s, a)-rectangularity, Break Curse=No2026.05 | 4 | |
| Robust-Q-FTRLRobust MG Model=Tabular, Uncertainty Set=fictitious (s, a_i)-rectangularity, Break Curse=Yes2026.05 | 6 | |
| L-Robust-Q-FTRLRobust MG Model=Linear, d = S max_{i∈[n]} A_i, Uncertainty Set=fictitious (s, a_i)-rectangularity, Break Curse=Yes2026.05 | 9 |