paper-with-me

Papers

Navigating the Designs of Privacy-Preserving Fine-tuning for Large Language Models

2025-01-08 · Haonan Shi, Tu Ouyang, An Wang

Instruction tuning has proven effective in enhancing Large Language Models' (LLMs) performance on downstream tasks. However, real-world fine-tuning faces inherent conflicts between model providers' intellectual property protection, clients' data privacy requirements, and tuning costs. While recent approaches like split learning and offsite tuning demonstrate promising architectures for privacy-preserving fine-tuning, there is a gap in systematically addressing the multidimensional trade-offs required for diverse real-world deployments. We propose several indicative evaluation metrics to guide design trade-offs for privacy-preserving fine-tuning and a series of example designs, collectively named GuardedTuning; they result from novel combinations of system architectures with adapted privacy-enhancement methods and emerging computation techniques. Each design represents distinct trade-offs across model utility, privacy guarantees, and costs. Experimental results demonstrate that these designs protect against data reconstruction attacks while maintaining competitive fine-tuning performance.

📄 PDF Abstract BibTeX arXiv:2501.04323

Code (0)

등록된 구현이 없습니다.

Tasks

Privacy Preserving

Similar Papers 제목 키워드 기반

Efficient Federated Prompt Tuning for Black-box Large Pre-trained Models

2023-10-04 · Zihao Lin, Yan Sun, Yifan Shi, Xueqian Wang 외

With the blowout development of pre-trained models (PTMs), the efficient tuning of these models for diverse downstream applications has emerged as a pivotal research concern. Although recent investigations into prompt tu…

I can't see it but I can Fine-tune it: On Encrypted Fine-tuning of Transformers using Fully Homomorphic Encryption

2024-02-14 · Prajwal Panzade, Daniel Takabi, Zhipeng Cai

In today's machine learning landscape, fine-tuning pretrained transformer models has emerged as an essential technique, particularly in scenarios where access to task-aligned training data is limited. However, challenges…

image-classificationImage ClassificationPrivacy Preserving

GradOT: Training-free Gradient-preserving Offsite-tuning for Large Language Models

2025-07-06 · Kai Yao, Zhaorui Tan, Penglei Gao, Lichun Li 외 arxiv

The rapid growth of large language models (LLMs) with traditional centralized fine-tuning emerges as a key technique for adapting these models to domain-specific challenges, yielding privacy risks for both model and data…

Offsite-Tuning: Transfer Learning without Full Model

2023-02-09 · Guangxuan Xiao, Ji Lin, Song Han

Transfer learning is important for foundation models to adapt to downstream tasks. However, many foundation models are proprietary, so users must share their data with model owners to fine-tune the models, which is costl…

modelPrivacy PreservingTransfer Learning

Synthesizing Privacy-Preserving Text Data via Finetuning without Finetuning Billion-Scale LLMs

2025-03-16 · Bowen Tan, Zheng Xu, Eric Xing, Zhiting Hu 외

Synthetic data offers a promising path to train models while preserving data privacy. Differentially private (DP) finetuning of large language models (LLMs) as data generator is effective, but is impractical when computa…

ClusteringPrivacy PreservingPrompt Engineering