Physical Reasoning on PHYRE Within-template 1.0
85.49Success Rate (AUCCESS)RPIN
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| RPINMechanism=LfD, Input=Initial scenes, bboxs, Supervision=Bboxs, masks, outcomes2022.10 | 85.49 | — | |
| Dec [Joint]Conditioning frames (tau)=1, Roll-out duration (tau')=102020.06 | 80 | 84.1 | |
| DQNConditioning frames (tau)=12020.06 | 77.6 | 81.4 | |
| Dec [Joint]Conditioning frames (tau)=1, Roll-out duration (tau')=02020.06 | 76.7 | 80.7 | |
| RANDConditioning frames (tau)=12020.06 | 13.7 | 7.7 | |
| MEMConditioning frames (tau)=12020.06 | 2.4 | 2.7 |