Downstream model-based control on Hopper OpenAI Gym (test)
2,253.51Accumulated RewardOurs
Evaluation results
| Method | Links | |
|---|---|---|
| OursPretrain=true2026.03 | 2,253.51 | |
| OursPretrain=false2026.03 | 554.92 | |
| TrajworldPretrain=true2026.03 | 534.32 | |
| TrajworldPretrain=false2026.03 | 366.26 | |
| TDMPretrain=false2026.03 | 242.34 | |
| MLPEnsemblePretrain=true2026.03 | 200.56 | |
| TDMPretrain=true2026.03 | 154.64 | |
| MLPEnsemblePretrain=false2026.03 | 147.37 |