• search
    search
  • reviewers
    reviewers
  • feeds
    feeds
  • assignments
    assignments
  • settings
  • logout

Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modifications.

Boyi WeiKaixuan HuangYangsibo HuangTinghao XieXiangyu QiMengzhou XiaPrateek MittalMengdi WangPeter Henderson
Published in: CoRR (2024)
Keyphrases