| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Medical Visual Question Answering | PathVQA | Overall Accuracy72.61 | 86 | |
| Medical Visual Question Answering | PathVQA (test) | Accuracy78.2 | 55 | |
| Medical Visual Question Answering | PathVQA | Accuracy76.8 | 50 | |
| Medical Visual Question Answering | PathVQA closed-end | Accuracy93.63 | 35 | |
| Vision-Language Medical Reasoning | PathVQA | Token Cost (tokens/question)0.7 | 29 | |
| Medical Visual Question Answering | PathVQA Open | Accuracy38.65 | 22 | |
| Hallucination detection | PathVQA | AUC82 | 20 | |
| Visual Question Answering | PathVQA | Accuracy (Closed)92.9 | 19 | |
| Visual Question Answering | PathVQA (test) | Overall Accuracy92.7 | 19 | |
| Medical Visual Question Answering (Free-text) | PathVQA OOD | Accuracy62.3 | 12 | |
| Multi-modal Question Answering | PathVQA | Accuracy65.9 | 12 | |
| Visual Question Answering | PathVQA | Accuracy74.4 | 6 | |
| Medical Visual Question Answering | PathVQA (held-out) | Accuracy59.9 | 6 | |
| Visual Question Answering | PathVQA | BLEU-163.93 | 5 | |
| Out-of-distribution Detection | PathVQA (PVQA) (test) | FPR6.24 | 5 |