| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Visual Question Answering | IconQA | Top-1 Acc54.91 | 19 | |
| Icon Question Answering | IconQA (test) | Accuracy (Img)95.69 | 13 | |
| Visual Question Answering | IconQA text-based multiple-choice downstream (test) | TextVQA Accuracy58.3 | 7 | |
| Visual Question Answering | IconQA img | Accuracy60.85 | 3 |