Improve dataset card: Add metadata, paper/code links, and sample usage

#1
by nielsr HF Staff - opened

This pull request significantly enhances the dataset card for the AIxCC Challenge Benchmark (C Challenges) by:

  • Adding task_categories (text-classification, text-generation) and tags (vulnerability-detection, patching, cybersecurity, llm, benchmark, aixcc, c-language) to the metadata for improved discoverability.
  • Specifying the language as en in the metadata.
  • Including a prominent link to the Hugging Face paper page at the top of the card.
  • Adding a prominent link to the associated GitHub repository at the top of the card.
  • Incorporating a detailed "Sample Usage" section with code snippets directly from the GitHub README, demonstrating how to test the Cyber Reasoning System (CRS) that utilizes this benchmark.
  • Updating the "Reference Implementation" section to remove the redundant GitHub link, while retaining the arXiv paper link as per guidelines.
  • The license statement in the content remains unchanged, as a specific machine-readable license identifier for the dataset was not provided.
Ready to merge
This branch is ready to get merged automatically.

Sign up or log in to comment