Forgetting-MarI: LLM Unlearning via Marginal Information Regularization
This addresses privacy and regulatory compliance for AI systems, particularly LLMs, by enabling controlled unlearning without retraining, though it is an incremental improvement over existing unlearning methods.
The paper tackles the problem of selectively removing specific data from trained large language models (LLMs) for privacy and compliance, introducing Forgetting-MarI, which outperforms state-of-the-art methods by reliably forgetting target data while better preserving general model performance across benchmarks.
As AI models are trained on ever-expanding datasets, the ability to remove the influence of specific data from trained models has become essential for privacy protection and regulatory compliance. Unlearning addresses this challenge by selectively removing parametric knowledge from the trained models without retraining from scratch, which is critical for resource-intensive models such as Large Language Models (LLMs). Existing unlearning methods often degrade model performance by removing more information than necessary when attempting to ''forget'' specific data. We introduce Forgetting-MarI, an LLM unlearning framework that provably removes only the additional (marginal) information contributed by the data to be unlearned, while preserving the information supported by the data to be retained. By penalizing marginal information, our method yields an explicit upper bound on the unlearn dataset's residual influence in the trained models, providing provable undetectability. Extensive experiments confirm that our approach outperforms current state-of-the-art unlearning methods, delivering reliable forgetting and better preserved general model performance across diverse benchmarks. This advancement represents an important step toward making AI systems more controllable and compliant with privacy and copyright regulations without compromising their effectiveness.