Novaspree

/

tofu-Gemma3-adapter-1

like 1

Hi! I'm Niels from the Hugging Face community science team.

This PR improves the model card for this LoRA adapter. Specifically, it:

• Links the model to the original paper: MAAT: Multi-phase Adapter-Aware Targeted Unlearning.
• Adds a link to the GitHub repository.
• Updates the YAML metadata with the correct base_model (google/gemma-3-4b-it), license (apache-2.0), and pipeline_tag (text-generation).
• Provides a summary of the MAAT framework and the 5WBENCH benchmark used for evaluation.

Feel free to merge this if it looks good!