# Visual Reasoning on GQA balanced (test-dev)

Source: [Visual Reasoning on GQA balanced (test-dev)](<https://www.sota2.com/research/sota/visual-reasoning-on-gqa-balanced-test-dev>)

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

Data updated: 2026-02-26T05:31:51.937Z

Evaluation of model performance on visual reasoning tasks using the balanced test-development split of the GQA dataset, measured by 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 | Visual Reasoning |
| Dataset (including any stated split) | GQA balanced (test-dev) |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 6 of 6 |
| 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-reasoning-on-gqa-balanced-test-dev.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| EVPG | 58.9 | Not recorded | 694238fb4561469362850508 |
| InternVL1.5 | 56.3 | Not recorded | 694238fb4561469362850508 |
| EVPG | 54.7 | Not recorded | 694238fb4561469362850508 |
| BLIP\_lavis | 51 | Not recorded | 694238fb4561469362850508 |
| SDVP | 50.4 | Not recorded | 694238fb4561469362850508 |
| BLIP\_hg | 48.6 | Not recorded | 694238fb4561469362850508 |

### EVPG

Accuracy: 58.9. Result ID: 69983f0d5698665ec49b5728.

[Reporting paper](<https://www.sota2.com/research/paper/694238fb4561469362850508.md>)

[Method origin paper](<https://www.sota2.com/research/paper/694238fb4561469362850508.md>)

Recorded evaluation configuration: {"Framework":"VisProg","Supervision Type":"Supervised","Backbone":"InternVL1.5"}. This record does not assert independent reproduction.

### InternVL1.5

Accuracy: 56.3. Result ID: 69983f0d5698665ec49b5725.

[Reporting paper](<https://www.sota2.com/research/paper/694238fb4561469362850508.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73081e209131cea59f8b.md>)

[Implementation](<https://github.com/OpenGVLab/InternVL>)

Recorded evaluation configuration: {"Framework":"VisProg","Supervision Type":"Zero-shot (baseline)"}. This record does not assert independent reproduction.

### EVPG

Accuracy: 54.7. Result ID: 69983f0d5698665ec49b5727.

[Reporting paper](<https://www.sota2.com/research/paper/694238fb4561469362850508.md>)

[Method origin paper](<https://www.sota2.com/research/paper/694238fb4561469362850508.md>)

Recorded evaluation configuration: {"Framework":"VisProg","Supervision Type":"Supervised","Backbone":"BLIP\_lavis"}. This record does not assert independent reproduction.

### BLIP\_lavis

Accuracy: 51. Result ID: 69983f0d5698665ec49b5724.

[Reporting paper](<https://www.sota2.com/research/paper/694238fb4561469362850508.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6bd81e209131cea30874.md>)

[Implementation](<https://github.com/salesforce/BLIP>)

Recorded evaluation configuration: {"Framework":"VisProg","Supervision Type":"Zero-shot (baseline)"}. This record does not assert independent reproduction.

### SDVP

Accuracy: 50.4. Result ID: 69983f0d5698665ec49b5726.

[Reporting paper](<https://www.sota2.com/research/paper/694238fb4561469362850508.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa771e1e209131cea71290.md>)

Recorded evaluation configuration: {"Framework":"VisProg","Supervision Type":"Supervised","Backbone":"BLIP\_hg"}. This record does not assert independent reproduction.

### BLIP\_hg

Accuracy: 48.6. Result ID: 69983f0d5698665ec49b5723.

[Reporting paper](<https://www.sota2.com/research/paper/694238fb4561469362850508.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6bd81e209131cea30874.md>)

[Implementation](<https://github.com/salesforce/BLIP>)

Recorded evaluation configuration: {"Framework":"VisProg","Supervision Type":"Zero-shot (baseline)"}. This record does not assert independent reproduction.

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