Verify evals on Papers with Code

Hi,

Niels here from the open-source team at Hugging Face. Congratulations on your work!

I've made the [paper](https://paperswithcode.co/paper/2210.03347) and [9 paper-native evaluations](https://paperswithcode.co/paper/2210.03347#results) available on Papers with Code.

The paper has results on [Image Understanding](https://paperswithcode.co/tasks/image-understanding) and [Document Understanding](https://paperswithcode.co/tasks/document-understanding) task pages.

The Pix2Struct Large results currently rank first on [RefExp](https://paperswithcode.co/benchmark/refexp?task=image-understanding&eval=8588) and [Screen2Words](https://paperswithcode.co/benchmark/screen2words?task=image-understanding&eval=8590).

The Pix2Struct Large result currently ranks second on [Widget Captioning](https://paperswithcode.co/benchmark/widget-captioning?task=image-understanding&eval=8589).

The Pix2Struct Large result currently ranks third on [OCR-VQA](https://paperswithcode.co/benchmark/ocr-vqa?task=image-understanding&eval=8587).

Would it be possible to verify these results and let me know if any score, model name, benchmark protocol, or openness metadata should be corrected?

You can also edit the task, methods, project page, and GitHub URL directly from the paper page using your Hugging Face account.

If you'd like to showcase the results in your repository README, you can copy these live leaderboard badges (or use the “Copy PwC badge” button in the Results section):

[](https://paperswithcode.co/api/v1/papers/2210.03347/leaderboard-badge-link?eval=8588)
[](https://paperswithcode.co/api/v1/papers/2210.03347/leaderboard-badge-link?eval=8590)
[](https://paperswithcode.co/api/v1/papers/2210.03347/leaderboard-badge-link?eval=8589)
[](https://paperswithcode.co/api/v1/papers/2210.03347/leaderboard-badge-link?eval=8587)

Kind regards,

Niels

Start with the linked Papers with Code paper and its nine paper-native evaluations, then compare each listed score, model name, benchmark protocol, and openness metadata with the Pix2Struct paper information. Check the Image Understanding and Document Understanding task pages and report or apply any needed corrections. Done means all nine evaluations have been verified.

Written by the indexing model from the issue text.