cais/Zephyr_RMU

Zephyr RMU

Zephyr 7B model with hazardous knowledge about biosecurity and cybersecurity "unlearned" using Representation Misdirection for Unlearning (RMU). For more details, please check our paper.

Model sources

• Base model: zephyr-7B-beta
• Repository: https://github.com/centerforaisafety/wmdp
• Website: https://www.wmdp.ai/
• Corpora used for unlearning: https://huggingface.co/datasets/cais/wmdp-corpora

Performance

Zephyr RMU has been evaluated on WMDP, MMLU and MT-Bench. Higher accuracy on MMLU and MT-Bench, and lower accuracy on WMDP are preferred.

	WMDP-Bio	WMDP-Cyber	MMLU	MT-Bench
Zephyr 7B	63.7	44.0	58.1	7.33
Zephyr RMU	31.2	28.2	57.1	7.10

Citation

If you find this useful in your research, please consider citing our paper:

@misc{li2024wmdp,
      title={The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning}, 
      author={Nathaniel Li and Alexander Pan and Anjali Gopal and Summer Yue and Daniel Berrios and Alice Gatti and Justin D. Li and Ann-Kathrin Dombrowski and Shashwat Goel and Long Phan and Gabriel Mukobi and Nathan Helm-Burger and Rassin Lababidi and Lennart Justen and Andrew B. Liu and Michael Chen and Isabelle Barrass and Oliver Zhang and Xiaoyuan Zhu and Rishub Tamirisa and Bhrugu Bharathi and Adam Khoja and Zhenqi Zhao and Ariel Herbert-Voss and Cort B. Breuer and Sam Marks and Oam Patel and Andy Zou and Mantas Mazeika and Zifan Wang and Palash Oswal and Weiran Liu and Adam A. Hunt and Justin Tienken-Harder and Kevin Y. Shih and Kemper Talley and John Guan and Russell Kaplan and Ian Steneker and David Campbell and Brad Jokubaitis and Alex Levinson and Jean Wang and William Qian and Kallol Krishna Karmakar and Steven Basart and Stephen Fitz and Mindy Levine and Ponnurangam Kumaraguru and Uday Tupakula and Vijay Varadharajan and Yan Shoshitaishvili and Jimmy Ba and Kevin M. Esvelt and Alexandr Wang and Dan Hendrycks},
      year={2024},
      eprint={2403.03218},
      archivePrefix={arXiv},
      primaryClass={cs.LG}
}