Targeting Alignment: Extracting Safety Classifiers of Aligned LLMs

@article{ferrand_targeting_2025,
  title   = {Targeting Alignment: Extracting Safety Classifiers of Aligned {LLMs}},
  author  = {Noirot Ferrand, Jean-Charles and Beugin, Yohan and Pauley, Eric and Sheatsley, Ryan and McDaniel, Patrick},
  journal = {arXiv preprint arXiv:2501.16534},
  year    = {2025},
  doi     = {10.48550/arXiv.2501.16534}
}