Add model card for DAA CLIPTextModel component with metadata and sample usage

#1
by nielsr HF Staff - opened

This PR adds a comprehensive model card for this repository, which hosts a CLIPTextModel component associated with the "Dynamic Attention Analysis for Backdoor Detection in Text-to-Image Diffusion Models" paper.

The update includes:

  • Metadata:
    • pipeline_tag: text-to-image for discoverability.
    • library_name: transformers to indicate compatibility with the Hugging Face Transformers library, based on the config.json.
    • license: apache-2.0.
  • Content:
    • Links to the paper (huggingface.co/papers/2504.20518) and the GitHub repository (github.com/Robin-WZQ/DAA).
    • An overview of the Dynamic Attention Analysis (DAA) method, including images from the GitHub README.
    • Sample usage code snippets (command-line examples) directly from the GitHub README, demonstrating how to use the DAA detection scripts.
    • The official BibTeX citation.

Please review and merge if everything looks good.

RobinWZQ changed pull request status to merged

Sign up or log in to comment