Preference Prediction on Meta-World Push-Wall (train)
90Reward AccuracyRFP
Evaluation Results
| Method | Links | |
|---|---|---|
| RFPModel grouping=Multi-Task w/ Reasons, Number of seeds=32026.03 | 90 | |
| BT-MultiModel grouping=Multi-Task, Number of seeds=32026.03 | 89.3 | |
| ReCouPLe-ECModel grouping=Multi-Task w/ Reasons, Number of seeds=32026.03 | 84.3 | |
| ReCouPLe-ICModel grouping=Multi-Task w/ Reasons, Number of seeds=32026.03 | 82.3 |