CBP-Tuning: Efficient Local Customization for Black-box Large Language Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhao, Jiaxuan, Gu, Naibin, Feng, Yuchen, Liu, Xiyu, Fu, Peng, Lin, Zheng, Wang, Weiping
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912587854118912
author Zhao, Jiaxuan
Gu, Naibin
Feng, Yuchen
Liu, Xiyu
Fu, Peng
Lin, Zheng
Wang, Weiping
author_facet Zhao, Jiaxuan
Gu, Naibin
Feng, Yuchen
Liu, Xiyu
Fu, Peng
Lin, Zheng
Wang, Weiping
contents The high costs of customizing large language models (LLMs) fundamentally limit their adaptability to user-specific needs. Consequently, LLMs are increasingly offered as cloud-based services, a paradigm that introduces critical limitations: providers struggle to support personalized customization at scale, while users face privacy risks when exposing sensitive data. To address this dual challenge, we propose Customized Black-box Prompt Tuning (CBP-Tuning), a novel framework that facilitates efficient local customization while preserving bidirectional privacy. Specifically, we design a two-stage framework: (1) a prompt generator trained on the server-side to capture domain-specific and task-agnostic capabilities, and (2) user-side gradient-free optimization that tailors soft prompts for individual tasks. This approach eliminates the need for users to access model weights or upload private data, requiring only a single customized vector per task while achieving effective adaptation. Furthermore, the evaluation of CBP-Tuning in the commonsense reasoning, medical and financial domain settings demonstrates superior performance compared to baselines, showcasing its advantages in task-agnostic processing and privacy preservation.
format Preprint
id arxiv_https___arxiv_org_abs_2509_12112
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CBP-Tuning: Efficient Local Customization for Black-box Large Language Models
Zhao, Jiaxuan
Gu, Naibin
Feng, Yuchen
Liu, Xiyu
Fu, Peng
Lin, Zheng
Wang, Weiping
Computation and Language
The high costs of customizing large language models (LLMs) fundamentally limit their adaptability to user-specific needs. Consequently, LLMs are increasingly offered as cloud-based services, a paradigm that introduces critical limitations: providers struggle to support personalized customization at scale, while users face privacy risks when exposing sensitive data. To address this dual challenge, we propose Customized Black-box Prompt Tuning (CBP-Tuning), a novel framework that facilitates efficient local customization while preserving bidirectional privacy. Specifically, we design a two-stage framework: (1) a prompt generator trained on the server-side to capture domain-specific and task-agnostic capabilities, and (2) user-side gradient-free optimization that tailors soft prompts for individual tasks. This approach eliminates the need for users to access model weights or upload private data, requiring only a single customized vector per task while achieving effective adaptation. Furthermore, the evaluation of CBP-Tuning in the commonsense reasoning, medical and financial domain settings demonstrates superior performance compared to baselines, showcasing its advantages in task-agnostic processing and privacy preservation.
title CBP-Tuning: Efficient Local Customization for Black-box Large Language Models
topic Computation and Language
url https://arxiv.org/abs/2509.12112