HalfCheetah Control on D4RL Medium-Replay v0
55.1Normalized ScoreCOMBO
Evaluation Results
| Method | Links | |
|---|---|---|
| COMBO2021.02 | 55.1 | |
| MOPO2021.02 | 53.1 | |
| BRAC-v2021.02 | 47.7 | |
| CQL2021.02 | 46.2 | |
| BRAC-pimplementation=reported2021.06 | 45.6 | |
| CQLMethod Category=Best RL Baseline, Seeds=52022.10 | 45.5 | |
| BRAC-p2021.02 | 45.4 | |
| IQLMethod Category=Weighted BC, Seeds=52022.10 | 44.2 | |
| IQL2023.12 | 44.2 | |
| PORMethod Category=Conditioned BC, Seeds=52022.10 | 43.5 | |
| Fisher-BRCimplementation=author-provided (re-run)2021.06 | 43.3 | |
| TD3+BCalpha=2.52021.06 | 43.3 | |
| DTAMP2023.12 | 42.6 | |
| Diffuser2023.12 | 42.2 | |
| CQLimplementation=author-provided (re-run)2021.06 | 41.9 | |
| TT2023.12 | 41.9 | |
| 10%BCMethod Category=Weighted BC, Seeds=52022.10 | 40.6 | |
| MOREL2021.02 | 40.2 | |
| DD2023.12 | 39.3 | |
| BEAR2021.02 | 38.6 | |
| BC2021.02 | 38.4 | |
| One-step RLMethod Category=Weighted BC, Seeds=52022.10 | 38.1 | |
| RvSMethod Category=Conditioned BC, Seeds=52022.10 | 38 | |
| DTMethod Category=Conditioned BC, Seeds=52022.10 | 36.6 | |
| CQL2023.12 | 36.6 | |
| DT2023.12 | 36.6 | |
| BCimplementation=own2021.06 | 34.7 | |
| SAC-off2021.02 | -2.4 |