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/2404.07839) and [40 paper-native evaluations](https://paperswithcode.co/paper/2404.07839#results) available on Papers with Code.

The paper has results on [Reasoning](https://paperswithcode.co/tasks/reasoning), [World Knowledge](https://paperswithcode.co/tasks/world-knowledge), [Coding Agents](https://paperswithcode.co/tasks/coding-agents), and 3 additional task pages.

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.

Kind regards,

Niels

Open the linked Papers with Code paper and its six task pages, then compare the 40 paper-native evaluation entries with the paper's reported results. Check each score, model name, benchmark protocol, and openness metadata; done means reporting any discrepancies or confirming that the entries are accurate.

Written by the indexing model from the issue text.