Robotic Manipulation on OGBench scene-sparse online
99Success RateReBRAC
Evaluation results
| Method | Links | |
|---|---|---|
| ReBRACTemporal Difference (TD) steps=1-step, Policy classification=Gaussian, Base method=ReBRAC2025.07 | 99 | |
| BFNTemporal Difference (TD) steps=1-step, Policy classification=Flow, Base method=BFN2025.07 | 99 | |
| QC-FQLPolicy classification=Q-chunking, Base method=FQL2025.07 | 99 | |
| QCPolicy classification=Q-chunking, Base method=QC2025.07 | 99 | |
| QC-IFQLPolicy classification=Q-chunking, Base method=IFQL2025.07 | 98 | |
| BFN-nTemporal Difference (TD) steps=n-step, Policy classification=Flow, Base method=BFN2025.07 | 97 | |
| FQLTemporal Difference (TD) steps=1-step, Policy classification=Flow, Base method=FQL2025.07 | 95 | |
| RLPDTemporal Difference (TD) steps=1-step, Policy classification=From Scratch, Base method=RLPD2025.07 | 94 | |
| SUPE-GTTemporal Difference (TD) steps=1-step, Policy classification=From Scratch, Base method=SUPE-GT2025.07 | 92 | |
| IFQL-nTemporal Difference (TD) steps=n-step, Policy classification=Flow, Base method=IFQL2025.07 | 92 | |
| RLPD-ACTemporal Difference (TD) steps=1-step, Policy classification=From Scratch, Base method=RLPD-AC2025.07 | 91 | |
| IFQLTemporal Difference (TD) steps=1-step, Policy classification=Flow, Base method=IFQL2025.07 | 74 | |
| FQL-nTemporal Difference (TD) steps=n-step, Policy classification=Flow, Base method=FQL2025.07 | 70 | |
| IQLTemporal Difference (TD) steps=1-step, Policy classification=Gaussian, Base method=IQL2025.07 | 39 |