# Medical Visual Question Answering on SLAKE English

Source: [Medical Visual Question Answering on SLAKE English](<https://www.sota2.com/research/sota/medical-visual-question-answering-on-slake-english>)

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

Data updated: 2026-04-08T07:16:59.298Z

Evaluation of model performance on the SLAKE English dataset for Medical Visual Question Answering, reporting overall token F1, open-ended token recall, and closed-ended accuracy.

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) | SLAKE English |
| Metric | Closed-Ended Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 9 of 9 |
| 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

- [Closed-Ended Accuracy](<https://www.sota2.com/research/sota/medical-visual-question-answering-on-slake-english.md?metric=Closed-Ended%20Accuracy>): Higher is better; unit: PERCENT.

- [Overall Token F1](<https://www.sota2.com/research/sota/medical-visual-question-answering-on-slake-english.md?metric=Overall%20Token%20F1>): Higher is better; unit: PERCENT.

- [Open-Ended Token Recall](<https://www.sota2.com/research/sota/medical-visual-question-answering-on-slake-english.md?metric=Open-Ended%20Token%20Recall>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Closed-Ended Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| BiomedGPT-B | 89.9 | Not recorded | 68fa79ff1e209131cea808a4 |
| MedGemma 4B | 87.6 | Not recorded | 68fa79ff1e209131cea808a4 |
| LLaVA-Med (BioMedCLIP) | 86.8 | Not recorded | 68fa79ff1e209131cea808a4 |
| Med-Gemini | 84.6 | Not recorded | 68fa79ff1e209131cea808a4 |
| Gemini 1.5 Flash | 80.9 | Not recorded | 68fa79ff1e209131cea808a4 |
| Gemini 1.5 Pro | 78.6 | Not recorded | 68fa79ff1e209131cea808a4 |
| o1 | 76.3 | Not recorded | 68fa79ff1e209131cea808a4 |
| Gemma 2 27B | 64.5 | Not recorded | 68fa79ff1e209131cea808a4 |
| Gemma 2 4B | 53 | Not recorded | 68fa79ff1e209131cea808a4 |

### BiomedGPT-B

Closed-Ended Accuracy: 89.9. Result ID: 69d5ff9747f34ce924be19f7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79ff1e209131cea808a4.md>)

Recorded evaluation configuration: {"Model Category":"Small Models"}. This record does not assert independent reproduction.

### MedGemma 4B

Closed-Ended Accuracy: 87.6. Result ID: 69d5ff9747f34ce924be19f4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79ff1e209131cea808a4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79ff1e209131cea808a4.md>)

Recorded evaluation configuration: {"Model Category":"Small Models"}. This record does not assert independent reproduction.

### LLaVA-Med (BioMedCLIP)

Closed-Ended Accuracy: 86.8. Result ID: 69d5ff9747f34ce924be19f8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79ff1e209131cea808a4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f8e1e209131cea46356.md>)

[Implementation](<https://aka.ms/llava-med>)

Recorded evaluation configuration: {"Model Category":"Small Models"}. This record does not assert independent reproduction.

### Med-Gemini

Closed-Ended Accuracy: 84.6. Result ID: 69d5ff9747f34ce924be19f9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79ff1e209131cea808a4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa730a1e209131cea5a0f8.md>)

Recorded evaluation configuration: {"Model Category":"Large Models"}. This record does not assert independent reproduction.

### Gemini 1.5 Flash

Closed-Ended Accuracy: 80.9. Result ID: 69d5ff9747f34ce924be19fa.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79ff1e209131cea808a4.md>)

Recorded evaluation configuration: {"Model Category":"Large Models"}. This record does not assert independent reproduction.

### Gemini 1.5 Pro

Closed-Ended Accuracy: 78.6. Result ID: 69d5ff9747f34ce924be19fb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79ff1e209131cea808a4.md>)

Recorded evaluation configuration: {"Model Category":"Large Models"}. This record does not assert independent reproduction.

### o1

Closed-Ended Accuracy: 76.3. Result ID: 69d5ff9747f34ce924be19fc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79ff1e209131cea808a4.md>)

Recorded evaluation configuration: {"Model Category":"Large Models"}. This record does not assert independent reproduction.

### Gemma 2 27B

Closed-Ended Accuracy: 64.5. Result ID: 69d5ff9747f34ce924be19f6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79ff1e209131cea808a4.md>)

Recorded evaluation configuration: {"Model Category":"Small Models"}. This record does not assert independent reproduction.

### Gemma 2 4B

Closed-Ended Accuracy: 53. Result ID: 69d5ff9747f34ce924be19f5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79ff1e209131cea808a4.md>)

Recorded evaluation configuration: {"Model Category":"Small Models"}. This record does not assert independent reproduction.

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