作者位次 #2/12ACL 2026Poster· Natural Language Processing
MTA :A Merge-then-Adapt Framework for Personalized Large Language Models
Xiaopeng Li · Yuanjin Zheng · an Jiaotong University) · Wenlin Zhang · an Jiaotong University) · Yingyi Zhang · Haiying He · Mengyang Ma · Yiqi Wang · Maolin WANG · Xuetao Wei · an Jiaotong University)
摘要
Personalized Large Language Models (PLLMs) aim to align model outputs with individual user preferences, a crucial capability for user-centric applications. However, the prevalent approach of fine-tuning a separate module for each user faces two major limitations: (1) storage costs scale linearly...