Offline Reinforcement Learning on OGBench State
92.5Cube Single Task 1 Success RateInFOM
Evaluation results
| Method | Links | ||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| InFOM2025.06 | 92.5 | 78.4 | 56.4 | 91.5 | 70 | 29.3 | 12.5 | 11.6 | 0.3 | 2.8 | 97.8 | 15.6 | 43.5 | 1 | 0.3 | 24.2 | 14.5 | 26.3 | 12 | 12.3 | |
| ReBRACAugmentation=MBPO2025.06 | 77.8 | 94.2 | 93.1 | 88.7 | 87.8 | 2.7 | 0 | 0 | 0 | 0 | 35.3 | 5.6 | 2.4 | 2 | 0 | 16.9 | 0.2 | 5.1 | 0.4 | 0 | |
| ReBRAC2025.06 | 67.3 | 93.7 | 94.8 | 89.5 | 83.3 | 2.2 | 0 | 0 | 0 | 0 | 47.7 | 7.8 | 1.7 | 2.8 | 0 | 12.8 | 0.5 | 5 | 0.8 | 0 | |
| IQL2025.06 | 53 | 51.7 | 41.5 | 42.2 | 33.7 | 17.8 | 1.3 | 0.3 | 0 | 1.5 | 66.5 | 2.5 | 0.7 | 0.2 | 0 | 2.3 | 0.5 | 1 | 0.3 | 0.7 | |
| IQLModel=HILP2025.06 | 32.9 | 26.5 | 35.5 | 36.4 | 18.5 | 0.7 | 0 | 0 | 0 | 0 | 8.8 | 1.2 | 0 | 0 | 0 | 0.3 | 0.4 | 0.1 | 0.1 | 0.1 | |
| IQLModel=FB2025.06 | 17.7 | 16.7 | 16 | 18.7 | 14.2 | 0.2 | 0 | 0 | 0 | 0 | 12.3 | 1.5 | 0 | 0.2 | 0 | 0.2 | 0.2 | 0.2 | 0.2 | 0 | |
| TD InfoNCEConditioning=IS2025.06 | 13.8 | 8.5 | 11.7 | 3.3 | 4 | 6.7 | 0 | 0 | 0 | 0.2 | 21 | 1.7 | 0.5 | 0.7 | 0 | 0 | 0 | 0 | 0 | 0 | |
| CRLConditioning=IS2025.06 | 10.1 | 3.7 | 12.5 | 1.7 | 4.3 | 4.1 | 0 | 0 | 0 | 0.3 | 17.5 | 2.3 | 0.8 | 1.2 | 0.3 | 0 | 0.3 | 0.3 | 0 | 0.5 | |
| ReBRACEncoder=DINO2025.06 | 1.8 | 1.2 | 1.5 | 0.5 | 0.5 | 0 | 0 | 0 | 0 | 0 | 26.7 | 1.3 | 0.2 | 0.2 | 0 | 0.3 | 0 | 0 | 0 | 0 |