# Vision Understanding on Common Visual Understanding Benchmarks (GQA, MMB, MME, POPE, SEED, SQA, VQAv2)

Source: [Vision Understanding on Common Visual Understanding Benchmarks (GQA, MMB, MME, POPE, SEED, SQA, VQAv2)](<https://www.sota2.com/research/sota/vision-understanding-on-common-visual-understanding-benchmarks-gqa-mmb-mme-pope-seed-sqa-vqav2>)

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

Data updated: 2026-02-26T05:45:48.487Z

Evaluation of general vision understanding capabilities across a suite comprising GQA, MMBench, MME, POPE, SEED, SQA, and VQAv2.

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 | Vision Understanding |
| Dataset (including any stated split) | Common Visual Understanding Benchmarks (GQA, MMB, MME, POPE, SEED, SQA, VQAv2) |
| Metric | GQA Score |
| 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

- [GQA Score](<https://www.sota2.com/research/sota/vision-understanding-on-common-visual-understanding-benchmarks-gqa-mmb-mme-pope-seed-sqa-vqav2.md?metric=GQA%20Score>): Higher is better; unit: PERCENT.

- [MMB Score](<https://www.sota2.com/research/sota/vision-understanding-on-common-visual-understanding-benchmarks-gqa-mmb-mme-pope-seed-sqa-vqav2.md?metric=MMB%20Score>): Higher is better; unit: PERCENT.

- [MME Score](<https://www.sota2.com/research/sota/vision-understanding-on-common-visual-understanding-benchmarks-gqa-mmb-mme-pope-seed-sqa-vqav2.md?metric=MME%20Score>): Higher is better; unit: SCALAR.

- [POPE Score](<https://www.sota2.com/research/sota/vision-understanding-on-common-visual-understanding-benchmarks-gqa-mmb-mme-pope-seed-sqa-vqav2.md?metric=POPE%20Score>): Higher is better; unit: PERCENT.

- [SEED Score](<https://www.sota2.com/research/sota/vision-understanding-on-common-visual-understanding-benchmarks-gqa-mmb-mme-pope-seed-sqa-vqav2.md?metric=SEED%20Score>): Higher is better; unit: PERCENT.

- [SQA Score](<https://www.sota2.com/research/sota/vision-understanding-on-common-visual-understanding-benchmarks-gqa-mmb-mme-pope-seed-sqa-vqav2.md?metric=SQA%20Score>): Higher is better; unit: PERCENT.

- [VQAv2 Score](<https://www.sota2.com/research/sota/vision-understanding-on-common-visual-understanding-benchmarks-gqa-mmb-mme-pope-seed-sqa-vqav2.md?metric=VQAv2%20Score>): Higher is better; unit: PERCENT.

- [Average Score](<https://www.sota2.com/research/sota/vision-understanding-on-common-visual-understanding-benchmarks-gqa-mmb-mme-pope-seed-sqa-vqav2.md?metric=Average%20Score>): Higher is better; unit: PERCENT.

- [Compression Performance](<https://www.sota2.com/research/sota/vision-understanding-on-common-visual-understanding-benchmarks-gqa-mmb-mme-pope-seed-sqa-vqav2.md?metric=Compression%20Performance>): Higher is better; unit: PERCENT.

## Recorded results

| Method | GQA Score | Result date | Report paper |
| --- | --- | --- | --- |
| Upper Bound | 61.1 | Not recorded | 68fa773e1e209131cea71ed4 |
| VoCo-LLaMA | 57 | Not recorded | 68fa773e1e209131cea71ed4 |
| Avg. Pool + Linear | 52.9 | Not recorded | 68fa773e1e209131cea71ed4 |
| Q-Former | 51.1 | Not recorded | 68fa773e1e209131cea71ed4 |
| Lower Bound | 37.7 | Not recorded | 68fa773e1e209131cea71ed4 |

### Upper Bound

GQA Score: 61.1. Result ID: 69946f443a6c3148491d3852.

[Reporting paper](<https://www.sota2.com/research/paper/68fa773e1e209131cea71ed4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6fe61e209131cea47fbc.md>)

[Implementation](<https://github.com/shikras/shikra>)

Recorded evaluation configuration: {"Token":"576"}. This record does not assert independent reproduction.

### VoCo-LLaMA

GQA Score: 57. Result ID: 69946f443a6c3148491d3853.

[Reporting paper](<https://www.sota2.com/research/paper/68fa773e1e209131cea71ed4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa773e1e209131cea71ed4.md>)

Recorded evaluation configuration: {"Token":"1"}. This record does not assert independent reproduction.

### Avg. Pool + Linear

GQA Score: 52.9. Result ID: 69946f443a6c3148491d3854.

[Reporting paper](<https://www.sota2.com/research/paper/68fa773e1e209131cea71ed4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71461e209131cea50559.md>)

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

Recorded evaluation configuration: {"Token":"1"}. This record does not assert independent reproduction.

### Q-Former

GQA Score: 51.1. Result ID: 69946f443a6c3148491d3855.

[Reporting paper](<https://www.sota2.com/research/paper/68fa773e1e209131cea71ed4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6fb71e209131cea46fe6.md>)

[Implementation](<https://github.com/salesforce/LAVIS/tree/main/projects/blip2>)

Recorded evaluation configuration: {"Token":"1"}. This record does not assert independent reproduction.

### Lower Bound

GQA Score: 37.7. Result ID: 69946f443a6c3148491d3856.

[Reporting paper](<https://www.sota2.com/research/paper/68fa773e1e209131cea71ed4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6fe61e209131cea47fbc.md>)

[Implementation](<https://github.com/shikras/shikra>)

Recorded evaluation configuration: {"Token":"1"}. This record does not assert independent reproduction.

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