# Visual-centric Reasoning on MMVP

Source: [Visual-centric Reasoning on MMVP](<https://www.sota2.com/research/sota/visual-centric-reasoning-on-mmvp>)

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

Data updated: 2026-02-26T05:35:49.922Z

Evaluation of model performance on the MMVP dataset focusing on Visual-centric Reasoning capabilities, measured by Average Score.

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-centric Reasoning |
| Dataset (including any stated split) | MMVP |
| Metric | Average Score |
| Unit | SCALAR |
| 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

- [Average Score](<https://www.sota2.com/research/sota/visual-centric-reasoning-on-mmvp.md?metric=Average%20Score>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Average Score | Result date | Report paper |
| --- | --- | --- | --- |
| SAIL-L-NV2 | 28.9 | Not recorded | 68fa78231e209131cea76f2e |
| SAIL-L-NV2 | 28 | Not recorded | 68fa78231e209131cea76f2e |
| SAIL-B-NV2 | 24.4 | Not recorded | 68fa78231e209131cea76f2e |
| DreamLIP | 24 | Not recorded | 68fa78231e209131cea76f2e |
| SAIL-B-GTE | 23 | Not recorded | 68fa78231e209131cea76f2e |
| SAIL-L-GTE | 22.2 | Not recorded | 68fa78231e209131cea76f2e |
| CLIP-L | 20 | Not recorded | 68fa78231e209131cea76f2e |
| CLIP-B | 19.3 | Not recorded | 68fa78231e209131cea76f2e |
| SAIL-L-GTE | 17 | Not recorded | 68fa78231e209131cea76f2e |

### SAIL-L-NV2

Average Score: 28.9. Result ID: 69957de33badfaffe67574e6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78231e209131cea76f2e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78231e209131cea76f2e.md>)

[Implementation](<https://lezhang7.github.io/sail.github.io/>)

Recorded evaluation configuration: {"Pre-training Data":"23M Merged","Model Architecture":"ViT-L/14"}. This record does not assert independent reproduction.

### SAIL-L-NV2

Average Score: 28. Result ID: 69957de33badfaffe67574e5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78231e209131cea76f2e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78231e209131cea76f2e.md>)

[Implementation](<https://lezhang7.github.io/sail.github.io/>)

Recorded evaluation configuration: {"Pre-training Data":"CC12M","Model Architecture":"ViT-L/14"}. This record does not assert independent reproduction.

### SAIL-B-NV2

Average Score: 24.4. Result ID: 69957de33badfaffe67574e1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78231e209131cea76f2e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78231e209131cea76f2e.md>)

[Implementation](<https://lezhang7.github.io/sail.github.io/>)

Recorded evaluation configuration: {"Pre-training Data":"CC12M","Model Architecture":"ViT-B/16","Patch Size":"14"}. This record does not assert independent reproduction.

### DreamLIP

Average Score: 24. Result ID: 69957de33badfaffe67574df.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78231e209131cea76f2e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72b51e209131cea5814c.md>)

Recorded evaluation configuration: {"Pre-training Data":"CC12M","Model Architecture":"ViT-B/16"}. This record does not assert independent reproduction.

### SAIL-B-GTE

Average Score: 23. Result ID: 69957de33badfaffe67574e0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78231e209131cea76f2e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78231e209131cea76f2e.md>)

[Implementation](<https://lezhang7.github.io/sail.github.io/>)

Recorded evaluation configuration: {"Pre-training Data":"CC12M","Model Architecture":"ViT-B/16","Patch Size":"14"}. This record does not assert independent reproduction.

### SAIL-L-GTE

Average Score: 22.2. Result ID: 69957de33badfaffe67574e4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78231e209131cea76f2e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78231e209131cea76f2e.md>)

[Implementation](<https://lezhang7.github.io/sail.github.io/>)

Recorded evaluation configuration: {"Pre-training Data":"23M Merged","Model Architecture":"ViT-L/14"}. This record does not assert independent reproduction.

### CLIP-L

Average Score: 20. Result ID: 69957de33badfaffe67574e7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78231e209131cea76f2e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69581e209131cea21ad0.md>)

[Implementation](<https://github.com/OpenAI/CLIP>)

Recorded evaluation configuration: {"Pre-training Data":"LAION400M","Model Architecture":"ViT-L/14"}. This record does not assert independent reproduction.

### CLIP-B

Average Score: 19.3. Result ID: 69957de33badfaffe67574e2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78231e209131cea76f2e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72b51e209131cea5814c.md>)

Recorded evaluation configuration: {"Pre-training Data":"LAION400M","Model Architecture":"ViT-B/16"}. This record does not assert independent reproduction.

### SAIL-L-GTE

Average Score: 17. Result ID: 69957de33badfaffe67574e3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78231e209131cea76f2e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78231e209131cea76f2e.md>)

[Implementation](<https://lezhang7.github.io/sail.github.io/>)

Recorded evaluation configuration: {"Pre-training Data":"CC12M","Model Architecture":"ViT-L/14"}. This record does not assert independent reproduction.

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