| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| LLaVA Adv N=8 OneVision (full evaluation set) | Accuracy60.16 | 4 | 21d ago | ||
| LLaVA Adv N=4 OneVision (full evaluation set) | Accuracy74.32 | 4 | 21d ago | ||
| LLaVA Random N=8 OneVision (full evaluation set) | LAD | Accuracy94.92 | 4 | 21d ago | |
| LLaVA Random N=4 full OneVision (evaluation) | Accuracy99.04 | 4 | 21d ago |