What Makes and Breaks Safety Fine-tuning? A Mechanistic Study

Item #:

079017-2963

Details

DOI: https://doi.org/10.52202/079017-2963
Author(s): Puneet Dokania, Samyak Jain, Tom Joy, Ekdeep Lubana, Kemal Oksuz, Amartya Sanyal, Philip Torr
Pages: 93406-93478 (73 pages)
Format: PDF Paper Download
Conference: Advances in Neural Information Processing Systems 37
Date/Location: Held 10-15 December 2024, Vancouver, Canada.
Series: Advances in Neural Information Processing Systems 37
Publisher: Neural Information Processing Systems Foundation, Inc. (NeurIPS)

DOI: https://doi.org/10.52202/079017-2963
Author(s): Puneet Dokania, Samyak Jain, Tom Joy, Ekdeep Lubana, Kemal Oksuz, Amartya Sanyal, Philip Torr
Pages: 93406-93478 (73 pages)
Format: PDF Paper Download
Conference: Advances in Neural Information Processing Systems 37
Date/Location: Held 10-15 December 2024, Vancouver, Canada.
Series: Advances in Neural Information Processing Systems 37
Publisher: Neural Information Processing Systems Foundation, Inc. (NeurIPS)