Offline Reinforcement Learning on Walker2D Medium-Expert (Noise 0)
109.6Normalized ReturnIQL
Evaluation Results
| Method | Links | |
|---|---|---|
| IQLModel Category=Model-Free2026.02 | 109.6 | |
| CQLModel Category=Model-Free2026.02 | 108.8 | |
| TAPModel Category=Model-Based2026.02 | 105.32 | |
| I-TAPModel Category=Model-Based2026.02 | 101.2 | |
| L-MAPModel Category=Model-Based2026.02 | 100.38 | |
| 1R2RModel Category=Model-Based2026.02 | 73.18 | |
| DTModel Category=Model-Free2026.02 | 64.65 |