# Image-to-TikZ on GAOKAO-MM Math

Source: [Image-to-TikZ on GAOKAO-MM Math](<https://www.sota2.com/research/sota/image-to-tikz-on-gaokao-mm-math>)

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

Data updated: 2026-03-25T13:06:23.774Z

Evaluation of model performance on generating TikZ code from input images using the GAOKAO-MM Math dataset.

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 | Image-to-TikZ |
| Dataset (including any stated split) | GAOKAO-MM Math |
| Metric | CLIP Score |
| Unit | PERCENT |
| 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

- [CLIP Score](<https://www.sota2.com/research/sota/image-to-tikz-on-gaokao-mm-math.md?metric=CLIP%20Score>): Higher is better; unit: PERCENT.

- [FID](<https://www.sota2.com/research/sota/image-to-tikz-on-gaokao-mm-math.md?metric=FID>): Lower is better; unit: SCALAR.

- [CSR](<https://www.sota2.com/research/sota/image-to-tikz-on-gaokao-mm-math.md?metric=CSR>): Higher is better; unit: PERCENT.

## Recorded results

| Method | CLIP Score | Result date | Report paper |
| --- | --- | --- | --- |
| GeoTikzBridge-Base-38B | 90 | Not recorded | 69c36c000636e5bf4afc6464 |
| GeoTikzBridge-Base-8B | 89 | Not recorded | 69c36c000636e5bf4afc6464 |
| InternVL3-78B | 87.1 | Not recorded | 69c36c000636e5bf4afc6464 |
| Qwen2.5-VL-72B | 85.4 | Not recorded | 69c36c000636e5bf4afc6464 |
| FigCodifier-8B | 84.9 | Not recorded | 69c36c000636e5bf4afc6464 |
| GLM4.5-V-106B-A12B | 83.5 | Not recorded | 69c36c000636e5bf4afc6464 |
| InternVL3.5-38B | 83.4 | Not recorded | 69c36c000636e5bf4afc6464 |
| Qwen3-VL-30B-A3B | 82.6 | Not recorded | 69c36c000636e5bf4afc6464 |
| Qwen2.5-VL-32B | 80.8 | Not recorded | 69c36c000636e5bf4afc6464 |

### GeoTikzBridge-Base-38B

CLIP Score: 90. Result ID: 69c3dd8147f34ce924ad4119.

[Reporting paper](<https://www.sota2.com/research/paper/69c36c000636e5bf4afc6464.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c36c000636e5bf4afc6464.md>)

Recorded evaluation configuration: {"Parameters":"38B"}. This record does not assert independent reproduction.

### GeoTikzBridge-Base-8B

CLIP Score: 89. Result ID: 69c3dd8147f34ce924ad4118.

[Reporting paper](<https://www.sota2.com/research/paper/69c36c000636e5bf4afc6464.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c36c000636e5bf4afc6464.md>)

Recorded evaluation configuration: {"Parameters":"8B"}. This record does not assert independent reproduction.

### InternVL3-78B

CLIP Score: 87.1. Result ID: 69c3dd8147f34ce924ad4114.

[Reporting paper](<https://www.sota2.com/research/paper/69c36c000636e5bf4afc6464.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78241e209131cea76fa5.md>)

Recorded evaluation configuration: {"Parameters":"78B"}. This record does not assert independent reproduction.

### Qwen2.5-VL-72B

CLIP Score: 85.4. Result ID: 69c3dd8147f34ce924ad4112.

[Reporting paper](<https://www.sota2.com/research/paper/69c36c000636e5bf4afc6464.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa773a1e209131cea71d3c.md>)

Recorded evaluation configuration: {"Parameters":"72B"}. This record does not assert independent reproduction.

### FigCodifier-8B

CLIP Score: 84.9. Result ID: 69c3dd8147f34ce924ad4117.

[Reporting paper](<https://www.sota2.com/research/paper/69c36c000636e5bf4afc6464.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa789d1e209131cea795a9.md>)

[Implementation](<https://github.com/mathllm/MathCoder>)

Recorded evaluation configuration: {"Parameters":"8B"}. This record does not assert independent reproduction.

### GLM4.5-V-106B-A12B

CLIP Score: 83.5. Result ID: 69c3dd8147f34ce924ad4116.

[Reporting paper](<https://www.sota2.com/research/paper/69c36c000636e5bf4afc6464.md>)

Recorded evaluation configuration: {"Parameters":"106B-A12B"}. This record does not assert independent reproduction.

### InternVL3.5-38B

CLIP Score: 83.4. Result ID: 69c3dd8147f34ce924ad4115.

[Reporting paper](<https://www.sota2.com/research/paper/69c36c000636e5bf4afc6464.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7af01e209131cea85dc9.md>)

Recorded evaluation configuration: {"Parameters":"38B"}. This record does not assert independent reproduction.

### Qwen3-VL-30B-A3B

CLIP Score: 82.6. Result ID: 69c3dd8147f34ce924ad4113.

[Reporting paper](<https://www.sota2.com/research/paper/69c36c000636e5bf4afc6464.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa788e1e209131cea79116.md>)

Recorded evaluation configuration: {"Parameters":"30B-A3B"}. This record does not assert independent reproduction.

### Qwen2.5-VL-32B

CLIP Score: 80.8. Result ID: 69c3dd8147f34ce924ad4111.

[Reporting paper](<https://www.sota2.com/research/paper/69c36c000636e5bf4afc6464.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa773a1e209131cea71d3c.md>)

Recorded evaluation configuration: {"Parameters":"32B"}. This record does not assert independent reproduction.

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