Goal-conditioned visual planning on COIN T=3 71
34.85Success Rate (SR)GeoWorld
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GeoWorldBackbone=ViT-g3842026.02 | 34.85 | 61.86 | 89.88 | |
| GeoWorldBackbone=ViT-g2026.02 | 34.41 | 60.47 | 89 | |
| V-JEPA 2Backbone=ViT-g3842026.02 | 34.08 | 57.2 | 89.53 | |
| GeoWorldBackbone=ViT-H2026.02 | 34.08 | 58.7 | 88.42 | |
| V-JEPA 2Backbone=ViT-g2026.02 | 33.42 | 56.29 | 87.31 | |
| GeoWorldBackbone=ViT-L2026.02 | 33.42 | 57.26 | 88.03 | |
| V-JEPA 2Backbone=ViT-H2026.02 | 32.76 | 55.37 | 86.57 | |
| V-JEPA 2Backbone=ViT-L2026.02 | 32.1 | 54.25 | 85.18 | |
| SCHEMA2026.02 | 32.09 | 49.84 | 83.83 | |
| LFP2026.02 | 30.64 | 54.72 | 76.86 | |
| MTID2026.02 | 30.44 | 51.7 | 59.74 | |
| VidAssistzero-shot=false2026.02 | 29.2 | 54.76 | 78.02 | |
| ActionDiffusion2026.02 | 24 | 45.42 | 54.29 | |
| PDPP2026.02 | 21.33 | 45.62 | 51.82 | |
| KEPP2026.02 | 20.25 | 39.87 | 51.72 | |
| E3P2026.02 | 19.57 | 31.42 | 84.95 | |
| VidAssistzero-shot=true2026.02 | 18.44 | 50.63 | 75.64 | |
| P3IV2026.02 | 15.4 | 21.67 | 76.31 | |
| DDN2026.02 | 13.9 | 20.19 | 64.78 | |
| Random2026.02 | 0.01 | 0.01 | 2.47 |