Offline Reinforcement Learning on MuJoCo walker2d medium-replay D4RL
94.1Normalized ReturnQCS-R
Evaluation results
| Method | Links | |
|---|---|---|
| QCS-RMethod Category=Ours2024.02 | 94.1 | |
| TD3+BCMethod Category=Value-Based2024.02 | 81.8 | |
| TD3+BC2024.10 | 81.8 | |
| SQLMethod Category=Value-Based2024.02 | 81.2 | |
| CGDTMethod Category=Combined Method2024.02 | 78.1 | |
| CQL2024.10 | 77.3 | |
| ReBRAC2024.10 | 77.3 | |
| CQLMethod Category=Value-Based2024.02 | 77.2 | |
| DCMethod Category=RCSL2024.02 | 76.6 | |
| PORMethod Category=Combined Method2024.02 | 76.6 | |
| EDTMethod Category=Combined Method2024.02 | 74.8 | |
| IQLMethod Category=Value-Based2024.02 | 73.9 | |
| IQL2024.10 | 73.9 | |
| DTMethod Category=RCSL2024.02 | 66.6 | |
| OTR+IQL2024.10 | 66 | |
| RvS-RMethod Category=RCSL2024.02 | 60.6 | |
| QDTMethod Category=Combined Method2024.02 | 58.2 | |
| ACTMethod Category=Combined Method2024.02 | 56.1 | |
| One-RL2024.10 | 49.5 | |
| BC2024.10 | 26 |