LUNE: Efficient LLM Unlearning via LoRA Fine-Tuning with Negative Examples
This addresses the need for practical unlearning in LLMs to handle privacy and bias issues, though it is incremental as it builds on existing LoRA methods.
The paper tackles the problem of efficiently removing specific information from large language models (LLMs) for privacy, bias mitigation, and knowledge correction, achieving effectiveness comparable to full fine-tuning while reducing computational cost by about an order of magnitude.
Large language models (LLMs) possess vast knowledge acquired from extensive training corpora, but they often cannot remove specific pieces of information when needed, which makes it hard to handle privacy, bias mitigation, and knowledge correction. Traditional model unlearning approaches require computationally expensive fine-tuning or direct weight editing, making them impractical for real-world deployment. In this work, we introduce LoRA-based Unlearning with Negative Examples (LUNE), a lightweight framework that performs negative-only unlearning by updating only low-rank adapters while freezing the backbone, thereby localizing edits and avoiding disruptive global changes. Leveraging Low-Rank Adaptation (LoRA), LUNE targets intermediate representations to suppress (or replace) requested knowledge with an order-of-magnitude lower compute and memory than full fine-tuning or direct weight editing. Extensive experiments on multiple factual unlearning tasks show that LUNE: (I) achieves effectiveness comparable to full fine-tuning and memory-editing methods, and (II) reduces computational cost by about an order of magnitude.