• 2024-07-06BADEDIT: BACKDOORING LARGE LANGUAGE MODELS BY MODEL EDITING
    本文是LLM系列文章,针对《BADEDIT:BACKDOORINGLARGELANGUAGEMODELSBYMODELEDITING》的翻译。BADEDIT:通过模型编辑后门攻击大型语言模型摘要1引言2背景和相关工作3后门攻击的轻量级编辑4BADEDIT5实验6结论摘要主流后门攻击方法通常需要大量的中