作者位次 #8/10ACL 2026Poster· Natural Language Processing
Preference Heads in Large Language Models: A Mechanistic Framework for Interpretable Personalization
Weixu Zhang · Ye Yuan · Changjiang Han · Yuxing Tian · Zipeng Sun · Linfeng Du · Jikun Kang · Hong Kang · Xue Liu · Haolun Wu
摘要
Large Language Models (LLMs) exhibit strong implicit personalization ability, yet most existing approaches treat this behavior as a black box, relying on prompt engineering or fine tuning on user data. In this work, we adopt a mechanistic interpretability perspective and hypothesize the existence...