Visual Reinforcement Learning on DMControl Cartpole, Swingup
872Episode ReturnCOVR
Evaluation Results
| Method | Links | |
|---|---|---|
| COVRType=VLM-assisted visual RL2026.01 | 872 | |
| PSRLType=Vanilla visual RL2026.01 | 849 | |
| APLType=VLM-assisted visual RL2026.01 | 820 | |
| ResActType=Vanilla visual RL2026.01 | 819 | |
| PlayVirtualType=Vanilla visual RL2026.01 | 816 | |
| MLRType=Vanilla visual RL2026.01 | 806 | |
| TACOType=Vanilla visual RL2026.01 | 782 | |
| DPLType=VLM-assisted visual RL2026.01 | 776 | |
| DrQType=Vanilla visual RL2026.01 | 759 | |
| VPFType=VLM-assisted visual RL2026.01 | 749 | |
| SVEAType=Vanilla visual RL2026.01 | 727 | |
| MADIType=Vanilla visual RL2026.01 | 704 | |
| CURLType=Vanilla visual RL2026.01 | 582 | |
| PlaNetType=Vanilla visual RL2026.01 | 563 | |
| DreamerType=Vanilla visual RL2026.01 | 326 | |
| VBEType=Only the VLM2026.01 | 178 |