| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| WorldPrediction-WM | VL-JEPA SFT | Accuracy65.7 | 20 | 4d ago | |
| ScienceWorld | NeSyS | Matter Score52.8 | 20 | 4d ago | |
| Plancraft | NeSyS | Smelt98.4 | 20 | 4d ago | |
| Webshop (test) | Search100 | 20 | 4d ago | ||
| MIND First 50 (test) | Matrix-Game 2.0 | Context Memory11.88 | 3 | 4d ago | |
| MIND-Third 50 1.0 (test) | Matrix-Game 2.0 | Long Context Memory0.1404 | 3 | 4d ago |