Online Reinforcement Learning on HopperStand DMControl (final)
874.63Normalized ReturnGoRL(Diff)
Evaluation results
| Method | Links | |
|---|---|---|
| GoRL(Diff)Seeds=5, Decoder=Diffusion2025.12 | 874.63 | |
| GoRL(FM)Seeds=5, Decoder=Flow-Matching2025.12 | 733.66 | |
| PPOSeeds=52025.12 | 286.09 | |
| FPOSeeds=52025.12 | 3.94 | |
| DPPOSeeds=52025.12 | 2.14 |