# Multimodal Generation on ART500K (test)

Source: [Multimodal Generation on ART500K (test)](<https://www.sota2.com/research/sota/multimodal-generation-on-art500k-test>)

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

Data updated: 2026-02-26T05:38:28.095Z

Evaluation of Multimodal Generation performance on the test split of the ART500K dataset, utilizing metrics such as CLIP scores (I2, IT), Perplexity, and KL Divergence.

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 | Multimodal Generation |
| Dataset (including any stated split) | ART500K (test) |
| Metric | CLIP I2 Score |
| Unit | SCALAR |
| 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

- [CLIP I2 Score](<https://www.sota2.com/research/sota/multimodal-generation-on-art500k-test.md?metric=CLIP%20I2%20Score>): Higher is better; unit: SCALAR.

- [Perplexity](<https://www.sota2.com/research/sota/multimodal-generation-on-art500k-test.md?metric=Perplexity>): Lower is better; unit: SCALAR.

- [CLIP IT Score](<https://www.sota2.com/research/sota/multimodal-generation-on-art500k-test.md?metric=CLIP%20IT%20Score>): Higher is better; unit: SCALAR.

- [KL Divergence](<https://www.sota2.com/research/sota/multimodal-generation-on-art500k-test.md?metric=KL%20Divergence>): Lower is better; unit: SCALAR.

## Recorded results

| Method | CLIP I2 Score | Result date | Report paper |
| --- | --- | --- | --- |
| GRAPHGPT-O | 77.62 | Not recorded | 68fa775d1e209131cea72c0c |
| GRAPHGPT-O | 72.64 | Not recorded | 68fa775d1e209131cea72c0c |
| DreamLLM | 71.15 | Not recorded | 68fa775d1e209131cea72c0c |
| Emu3 | 62.11 | Not recorded | 68fa775d1e209131cea72c0c |
| Chameleon | 61.19 | Not recorded | 68fa775d1e209131cea72c0c |
| Janus | 59.32 | Not recorded | 68fa775d1e209131cea72c0c |

### GRAPHGPT-O

CLIP I2 Score: 77.62. Result ID: 699499833a6c3148491e504e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa775d1e209131cea72c0c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa775d1e209131cea72c0c.md>)

Recorded evaluation configuration: {"Type":"Hard"}. This record does not assert independent reproduction.

### GRAPHGPT-O

CLIP I2 Score: 72.64. Result ID: 699499833a6c3148491e504f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa775d1e209131cea72c0c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa775d1e209131cea72c0c.md>)

Recorded evaluation configuration: {"Type":"Soft"}. This record does not assert independent reproduction.

### DreamLLM

CLIP I2 Score: 71.15. Result ID: 699499833a6c3148491e504d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa775d1e209131cea72c0c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72941e209131cea5742f.md>)

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

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Emu3

CLIP I2 Score: 62.11. Result ID: 699499833a6c3148491e504b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa775d1e209131cea72c0c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa732a1e209131cea5add9.md>)

[Implementation](<https://github.com/baaivision/Emu>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Chameleon

CLIP I2 Score: 61.19. Result ID: 699499833a6c3148491e504c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa775d1e209131cea72c0c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77971e209131cea74348.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Janus

CLIP I2 Score: 59.32. Result ID: 699499833a6c3148491e504a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa775d1e209131cea72c0c.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

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