Saved in:
Bibliographic Details
Main Authors: Zhao, Ziyi, Gao, Chongming, Zhang, Yang, Liu, Haoyan, Gan, Weinan, Guo, Huifeng, Liu, Yong, Feng, Fuli
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2601.12034
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909993347842048
author Zhao, Ziyi
Gao, Chongming
Zhang, Yang
Liu, Haoyan
Gan, Weinan
Guo, Huifeng
Liu, Yong
Feng, Fuli
author_facet Zhao, Ziyi
Gao, Chongming
Zhang, Yang
Liu, Haoyan
Gan, Weinan
Guo, Huifeng
Liu, Yong
Feng, Fuli
contents Personalization in Large Language Models (LLMs) often relies on user-specific soft prompts. However, these prompts become obsolete when the foundation model is upgraded, necessitating costly, full-scale retraining. To overcome this limitation, we propose the Prompt-level User Migration Adapter (PUMA), a lightweight framework to efficiently migrate personalized prompts across incompatible models. PUMA utilizes a parameter-efficient adapter to bridge the semantic gap, combined with a group-based user selection strategy to significantly reduce training costs. Experiments on three large-scale datasets show our method matches or even surpasses the performance of retraining from scratch, reducing computational cost by up to 98%. The framework demonstrates strong generalization across diverse model architectures and robustness in advanced scenarios like chained and aggregated migrations, offering a practical path for the sustainable evolution of personalized AI by decoupling user assets from the underlying models.
format Preprint
id arxiv_https___arxiv_org_abs_2601_12034
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Don't Start Over: A Cost-Effective Framework for Migrating Personalized Prompts Between LLMs
Zhao, Ziyi
Gao, Chongming
Zhang, Yang
Liu, Haoyan
Gan, Weinan
Guo, Huifeng
Liu, Yong
Feng, Fuli
Computation and Language
Information Retrieval
Personalization in Large Language Models (LLMs) often relies on user-specific soft prompts. However, these prompts become obsolete when the foundation model is upgraded, necessitating costly, full-scale retraining. To overcome this limitation, we propose the Prompt-level User Migration Adapter (PUMA), a lightweight framework to efficiently migrate personalized prompts across incompatible models. PUMA utilizes a parameter-efficient adapter to bridge the semantic gap, combined with a group-based user selection strategy to significantly reduce training costs. Experiments on three large-scale datasets show our method matches or even surpasses the performance of retraining from scratch, reducing computational cost by up to 98%. The framework demonstrates strong generalization across diverse model architectures and robustness in advanced scenarios like chained and aggregated migrations, offering a practical path for the sustainable evolution of personalized AI by decoupling user assets from the underlying models.
title Don't Start Over: A Cost-Effective Framework for Migrating Personalized Prompts Between LLMs
topic Computation and Language
Information Retrieval
url https://arxiv.org/abs/2601.12034