Task Adaptation on cheetah-jump
54.4Avg Mean Undiscounted ReturnIQL+S2P
Evaluation Results
| Method | Links | |
|---|---|---|
| IQL+S2Pzero-shot=true2022.09 | 54.4 | |
| CQL+S2Pzero-shot=true2022.09 | 47.2 | |
| CQLzero-shot=true2022.09 | 40.8 | |
| IQLzero-shot=true2022.09 | 36.6 | |
| Batch Meanzero-shot=true2022.09 | 2.05 |