# Visual Question Answering on WSI

Source: [Visual Question Answering on WSI](<https://www.sota2.com/research/sota/visual-question-answering-on-wsi>)

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

Data updated: 2026-02-26T05:54:16.330Z

Evaluation of model performance on the WSI dataset for Visual Question Answering, reporting Open and Closed 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 | Visual Question Answering |
| Dataset (including any stated split) | WSI |
| Metric | Open Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 2 of 2 |
| 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

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

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

## Recorded results

| Method | Open Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| CPath-Omni | 67.3 | Not recorded | 68fa76321e209131cea6bc46 |
| GPT-4o | 20.5 | Not recorded | 68fa76321e209131cea6bc46 |

### CPath-Omni

Open Accuracy: 67.3. Result ID: 699415e13a6c3148491af032.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76321e209131cea6bc46.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76321e209131cea6bc46.md>)

Recorded evaluation configuration: {"type":"foundational multimodal"}. This record does not assert independent reproduction.

### GPT-4o

Open Accuracy: 20.5. Result ID: 699415e13a6c3148491af031.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76321e209131cea6bc46.md>)

Recorded evaluation configuration: {"type":"general-purpose","context":"generated WSI reports"}. This record does not assert independent reproduction.

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