Year: 2025
Type: conference-paper
Abstract: 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 realworld 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.
Authors: Haonan Shi, Tu Ouyang, An Wang
Institution: Case Western Reserve University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Awards: G2767311645, G7276038111, G8876996369
OpenAlex ID: W4411549693