# Visual Question Answering on GQA (Accuracy and Relative Accuracy)

Source: [Visual Question Answering on GQA (Accuracy and Relative Accuracy)](<https://www.sota2.com/research/sota/visual-question-answering-on-gqa-accuracy-and-relative-accuracy>)

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

Data updated: 2026-05-29T09:27:42.082Z

Evaluation of model performance on the GQA dataset for Visual Question Answering, reporting Accuracy and Relative Accuracy 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) | GQA |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher 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

- [Accuracy](<https://www.sota2.com/research/sota/visual-question-answering-on-gqa-accuracy-and-relative-accuracy.md?metric=Accuracy>): Higher is better; unit: PERCENT.

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

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Vanilla | 63.2 | Not recorded | 6a191d88dbd7150506f827cb |
| OccamToken | 61.2 | Not recorded | 6a191d88dbd7150506f827cb |
| OccamToken | 59.3 | Not recorded | 6a191d88dbd7150506f827cb |
| VisionZip | 57.9 | Not recorded | 6a191d88dbd7150506f827cb |
| VisionZip | 56.2 | Not recorded | 6a191d88dbd7150506f827cb |

### Vanilla

Accuracy: 63.2. Result ID: 6a195bfafbc373592c140608.

[Reporting paper](<https://www.sota2.com/research/paper/6a191d88dbd7150506f827cb.md>)

Recorded evaluation configuration: {"Backbone":"LLaVA-v1.5 13B","Token Retention Budget":"576","Token Retention Ratio":"100%"}. This record does not assert independent reproduction.

### OccamToken

Accuracy: 61.2. Result ID: 6a195bfafbc373592c14060a.

[Reporting paper](<https://www.sota2.com/research/paper/6a191d88dbd7150506f827cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a191d88dbd7150506f827cb.md>)

Recorded evaluation configuration: {"Backbone":"LLaVA-v1.5 13B","Token Retention Budget":"128","Token Retention Ratio":"22.2%"}. This record does not assert independent reproduction.

### OccamToken

Accuracy: 59.3. Result ID: 6a195bfafbc373592c14060c.

[Reporting paper](<https://www.sota2.com/research/paper/6a191d88dbd7150506f827cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a191d88dbd7150506f827cb.md>)

Recorded evaluation configuration: {"Backbone":"LLaVA-v1.5 13B","Token Retention Budget":"64","Token Retention Ratio":"11.1%"}. This record does not assert independent reproduction.

### VisionZip

Accuracy: 57.9. Result ID: 6a195bfafbc373592c140609.

[Reporting paper](<https://www.sota2.com/research/paper/6a191d88dbd7150506f827cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa760f1e209131cea6b33b.md>)

[Implementation](<https://github.com/dvlab-research/VisionZip>)

Recorded evaluation configuration: {"Backbone":"LLaVA-v1.5 13B","Token Retention Budget":"128","Token Retention Ratio":"22.2%"}. This record does not assert independent reproduction.

### VisionZip

Accuracy: 56.2. Result ID: 6a195bfafbc373592c14060b.

[Reporting paper](<https://www.sota2.com/research/paper/6a191d88dbd7150506f827cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa760f1e209131cea6b33b.md>)

[Implementation](<https://github.com/dvlab-research/VisionZip>)

Recorded evaluation configuration: {"Backbone":"LLaVA-v1.5 13B","Token Retention Budget":"64","Token Retention Ratio":"11.1%"}. This record does not assert independent reproduction.

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