# Medical Visual Question Answering on PMC-VQA (300 held-out samples)

Source: [Medical Visual Question Answering on PMC-VQA (300 held-out samples)](<https://www.sota2.com/research/sota/medical-visual-question-answering-on-pmc-vqa-300-held-out-samples>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-06-02T10:19:58.125Z

Evaluation of model performance on the specific 300 held-out samples subset of the PMC-VQA dataset for Medical Visual Question Answering, reporting Risk, Accuracy, and Rewrite metrics.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Medical Visual Question Answering |
| Dataset (including any stated split) | PMC-VQA (300 held-out samples) |
| Metric | Risk |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 5 of 5 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Risk](<https://www.sota2.com/research/sota/medical-visual-question-answering-on-pmc-vqa-300-held-out-samples.md?metric=Risk>): Lower is better; unit: SCALAR.

- [Accuracy](<https://www.sota2.com/research/sota/medical-visual-question-answering-on-pmc-vqa-300-held-out-samples.md?metric=Accuracy>): Higher is better; unit: PERCENT.

- [Rewrite](<https://www.sota2.com/research/sota/medical-visual-question-answering-on-pmc-vqa-300-held-out-samples.md?metric=Rewrite>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Risk | Result date | Report paper |
| --- | --- | --- | --- |
| Oracle Bound | 0.604 | Not recorded | 6a1e63fcdbd7150506f8555f |
| Risk Rerank | 0.62 | Not recorded | 6a1e63fcdbd7150506f8555f |
| Utility Score | 0.62 | Not recorded | 6a1e63fcdbd7150506f8555f |
| Original | 0.6483 | Not recorded | 6a1e63fcdbd7150506f8555f |
| Visual Rewrite | 0.659 | Not recorded | 6a1e63fcdbd7150506f8555f |

### Oracle Bound

Risk: 0.604. Result ID: 6a1eae35fbc373592c1976aa.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63fcdbd7150506f8555f.md>)

Recorded evaluation configuration: {"Strategy":"Oracle Bound"}. This record does not assert independent reproduction.

### Risk Rerank

Risk: 0.62. Result ID: 6a1eae35fbc373592c1976a8.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63fcdbd7150506f8555f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1e63fcdbd7150506f8555f.md>)

Recorded evaluation configuration: {"Strategy":"Risk Rerank"}. This record does not assert independent reproduction.

### Utility Score

Risk: 0.62. Result ID: 6a1eae35fbc373592c1976a9.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63fcdbd7150506f8555f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1e63fcdbd7150506f8555f.md>)

Recorded evaluation configuration: {"Strategy":"Utility Score"}. This record does not assert independent reproduction.

### Original

Risk: 0.6483. Result ID: 6a1eae35fbc373592c1976a6.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63fcdbd7150506f8555f.md>)

Recorded evaluation configuration: {"Strategy":"Original"}. This record does not assert independent reproduction.

### Visual Rewrite

Risk: 0.659. Result ID: 6a1eae35fbc373592c1976a7.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63fcdbd7150506f8555f.md>)

Recorded evaluation configuration: {"Strategy":"Visual Rewrite"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.