# Cross-modal relevance matching on Commercial Advertising Dataset (test)

Source: [Cross-modal relevance matching on Commercial Advertising Dataset (test)](<https://www.sota2.com/research/sota/cross-modal-relevance-matching-on-commercial-advertising-dataset-test>)

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

Data updated: 2026-02-26T05:24:36.400Z

Evaluation of Cross-modal relevance matching performance on the test split of the Commercial Advertising Dataset, measured using AUC.

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 | Cross-modal relevance matching |
| Dataset (including any stated split) | Commercial Advertising Dataset (test) |
| Metric | AUC |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 18 of 18 |
| 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

- [AUC](<https://www.sota2.com/research/sota/cross-modal-relevance-matching-on-commercial-advertising-dataset-test.md?metric=AUC>): Higher is better; unit: PERCENT.

## Recorded results

| Method | AUC | Result date | Report paper |
| --- | --- | --- | --- |
| AlignCMSS | 91.73 | Not recorded | 68fa70b61e209131cea4cbd5 |
| AlignCMSS | 91.02 | Not recorded | 68fa70b61e209131cea4cbd5 |
| OSCAR | 90.62 | Not recorded | 68fa70b61e209131cea4cbd5 |
| AlignCMSS | 90.46 | Not recorded | 68fa70b61e209131cea4cbd5 |
| VL-BERT | 90.13 | Not recorded | 68fa70b61e209131cea4cbd5 |
| AdsCVLR | 89.16 | Not recorded | 68fa70b61e209131cea4cbd5 |
| VinVL | 88.56 | Not recorded | 68fa70b61e209131cea4cbd5 |
| Unicoder-VL@(Qt-OtOi) | 87.9 | Not recorded | 68fa70b61e209131cea4cbd5 |
| OSCAR | 87.45 | Not recorded | 68fa70b61e209131cea4cbd5 |
| VL-BERT | 86.27 | Not recorded | 68fa70b61e209131cea4cbd5 |
| BLIP | 83.51 | Not recorded | 68fa70b61e209131cea4cbd5 |
| Unicoder-VL | 83.16 | Not recorded | 68fa70b61e209131cea4cbd5 |
| ALBEF | 82.74 | Not recorded | 68fa70b61e209131cea4cbd5 |
| ViLT | 82.23 | Not recorded | 68fa70b61e209131cea4cbd5 |
| CLIP-MH@(Qt-OtOi) | 82.18 | Not recorded | 68fa70b61e209131cea4cbd5 |
| BERT-base | 82.15 | Not recorded | 68fa70b61e209131cea4cbd5 |
| CLIP@(Qt-OtOi) | 81.98 | Not recorded | 68fa70b61e209131cea4cbd5 |
| CLIP | 81.16 | Not recorded | 68fa70b61e209131cea4cbd5 |

### AlignCMSS

AUC: 91.73. Result ID: 699155533a6c314849076167.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Implementation](<https://github.com/Pter61/AlignCMSS/>)

Recorded evaluation configuration: {"Ads Input":"Multi-Modal","Training Data Percentage":"100%"}. This record does not assert independent reproduction.

### AlignCMSS

AUC: 91.02. Result ID: 699155533a6c314849076166.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Implementation](<https://github.com/Pter61/AlignCMSS/>)

Recorded evaluation configuration: {"Ads Input":"Multi-Modal","Training Data Percentage":"70%"}. This record does not assert independent reproduction.

### OSCAR

AUC: 90.62. Result ID: 699155533a6c314849076164.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Implementation](<https://github.com/Pter61/AlignCMSS/>)

Recorded evaluation configuration: {"Ads Input":"Multi-Modal","VALSE Integration":true}. This record does not assert independent reproduction.

### AlignCMSS

AUC: 90.46. Result ID: 699155533a6c314849076165.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Implementation](<https://github.com/Pter61/AlignCMSS/>)

Recorded evaluation configuration: {"Ads Input":"Multi-Modal","Training Data Percentage":"50%"}. This record does not assert independent reproduction.

### VL-BERT

AUC: 90.13. Result ID: 699155533a6c314849076163.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Implementation](<https://github.com/Pter61/AlignCMSS/>)

Recorded evaluation configuration: {"Ads Input":"Multi-Modal","VALSE Integration":true}. This record does not assert independent reproduction.

### AdsCVLR

AUC: 89.16. Result ID: 699155533a6c314849076162.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

Recorded evaluation configuration: {"Ads Input":"Multi-Modal"}. This record does not assert independent reproduction.

### VinVL

AUC: 88.56. Result ID: 699155533a6c314849076161.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa697d1e209131cea22699.md>)

[Implementation](<https://github.com/pzzhang/VinVL>)

Recorded evaluation configuration: {"Ads Input":"Multi-Modal"}. This record does not assert independent reproduction.

### Unicoder-VL@(Qt-OtOi)

AUC: 87.9. Result ID: 699155533a6c314849076160.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

Recorded evaluation configuration: {"Ads Input":"Multi-Modal"}. This record does not assert independent reproduction.

### OSCAR

AUC: 87.45. Result ID: 699155533a6c31484907615f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Implementation](<https://github.com/Pter61/AlignCMSS/>)

Recorded evaluation configuration: {"Ads Input":"Multi-Modal","VALSE Integration":false}. This record does not assert independent reproduction.

### VL-BERT

AUC: 86.27. Result ID: 699155533a6c31484907615e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Implementation](<https://github.com/Pter61/AlignCMSS/>)

Recorded evaluation configuration: {"Ads Input":"Multi-Modal","VALSE Integration":false}. This record does not assert independent reproduction.

### BLIP

AUC: 83.51. Result ID: 699155533a6c31484907615d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

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

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

Recorded evaluation configuration: {"Ads Input":"Multi-Modal"}. This record does not assert independent reproduction.

### Unicoder-VL

AUC: 83.16. Result ID: 699155533a6c314849076159.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66cb1e209131cea1258f.md>)

Recorded evaluation configuration: {"Ads Input":"Image"}. This record does not assert independent reproduction.

### ALBEF

AUC: 82.74. Result ID: 699155533a6c31484907615c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6aef1e209131cea2b2da.md>)

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

Recorded evaluation configuration: {"Ads Input":"Multi-Modal"}. This record does not assert independent reproduction.

### ViLT

AUC: 82.23. Result ID: 699155533a6c314849076158.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a1e1e209131cea260db.md>)

[Implementation](<https://github.com/dandelin/vilt>)

Recorded evaluation configuration: {"Ads Input":"Image"}. This record does not assert independent reproduction.

### CLIP-MH@(Qt-OtOi)

AUC: 82.18. Result ID: 699155533a6c31484907615b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

Recorded evaluation configuration: {"Ads Input":"Multi-Modal"}. This record does not assert independent reproduction.

### BERT-base

AUC: 82.15. Result ID: 699155533a6c314849076156.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65c31e209131cea0c4ed.md>)

[Implementation](<https://github.com/google-research/bert>)

Recorded evaluation configuration: {"Ads Input":"Text"}. This record does not assert independent reproduction.

### CLIP@(Qt-OtOi)

AUC: 81.98. Result ID: 699155533a6c31484907615a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

Recorded evaluation configuration: {"Ads Input":"Multi-Modal"}. This record does not assert independent reproduction.

### CLIP

AUC: 81.16. Result ID: 699155533a6c314849076157.

[Reporting paper](<https://www.sota2.com/research/paper/68fa70b61e209131cea4cbd5.md>)

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

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

Recorded evaluation configuration: {"Ads Input":"Image"}. This record does not assert independent reproduction.

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