Wisdom is Knowing What not to Say: Hallucination-Free LLMs Unlearning via Attention Shifting
智慧在于知道什么不说:通过注意力转移实现无幻觉的LLM去学习
机构 * Faculty of Information Technology, Monash University(墨尔本大学信息科技学院) ; Key Laboratory of Computing Power Network and Information Security, Ministry of Education, Shandong Computer Science Center, Qilu University of Technology (Shandong Academy of Sciences)(教育部计算能力网络与信息安全重点实验室,山东计算机科学中心,齐鲁大学(山东科学院)) ; Shandong Provincial Key Laboratory of Computing Power Internet and Service Computing, Shandong Fundamental Research Center for Computer Science(山东省计算能力互联网与服务计算重点实验室,山东省计算机科学基础研究中心) ; School of Computer Science and Technology, Anhui University(安徽大学计算机科学与技术学院)
AI总结 本文提出注意力转移框架,通过保留上下文抑制和抗幻觉响应塑造,提升LLM去学习效果,在保持性能的同时减少幻觉响应。
Comments Accepted by NeurIPS 2025