| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Image Quality Assessment | QBench | Accuracy77.5 | 81 | |
| Multi-image understanding | QBench2 | Accuracy81.7 | 30 | |
| Multi-image reasoning | QBench2 (val) | Accuracy79.3 | 21 | |
| Image Quality Assessment | QBench (test) | Accuracy74.1 | 17 | |
| Visual Quality Assessment | QBench | QBench Score60.6 | 12 |