LLMs Encode Harmfulness and Refusal Separately

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Zhao, Jiachen, Huang, Jing, Wu, Zhengxuan, Bau, David, Shi, Weiyan
Format:	Preprint
Published:	2025
Subjects:	Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Internet

https://arxiv.org/abs/2507.11878