Offline Policy Optimization on D4RL v2
33.9Hopper Score (Random)PhyB
Evaluation Results
| Method | Links | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| PhyB2026.05 | 33.9 | 34.7 | 23.5 | 109.4 | 74.5 | 95.5 | 110.7 | 74.7 | 85.4 | 116.5 | 109.4 | 112.4 | 81.7 | |
| ADM2026.05 | 32.7 | 45.4 | 22.2 | 107.4 | 72.2 | 93.2 | 104.4 | 67.6 | 95.6 | 112.7 | 103.7 | 114.9 | 81 | |
| MOBILE2026.05 | 31.9 | 39.3 | 17.9 | 106.6 | 74.6 | 87.7 | 103.9 | 71.7 | 89.9 | 112.6 | 108.2 | 115.2 | 80 | |
| 1R2R2026.05 | 30.9 | 36.9 | 7.6 | 80.2 | 74.5 | 63.9 | 92.9 | 65.7 | 92.2 | 81.6 | 96 | 90.9 | 67.8 | |
| ORPO2026.05 | 9.2 | 40.8 | 10.8 | 30.4 | 73.4 | 55.5 | 104.6 | 72.8 | 91.1 | 111 | 101.5 | 108.8 | 67.5 |