Offline Reinforcement Learning on Hopper Medium-Expert Noise 5
0.8287Normalized ReturnI-TAP
Evaluation Results
| Method | Links | |
|---|---|---|
| I-TAPModel Category=Model-Based2026.02 | 0.8287 | |
| L-MAPModel Category=Model-Based2026.02 | 0.7149 | |
| CQLModel Category=Model-Free2026.02 | 0.6803 | |
| DTModel Category=Model-Free2026.02 | 0.6187 | |
| IQLModel Category=Model-Free2026.02 | 0.4483 | |
| 1R2RModel Category=Model-Based2026.02 | 0.3799 | |
| TAPModel Category=Model-Based2026.02 | 0.3731 |