<ul><li>The article discusses the task of removing sensitive personal information from trained machine learning models.</li><li>Proposed method focuses on identifying the subset of model parameters that have the largest contribution in the unlearning process.</li><li>The selection of these parameters and updating them during unlearning leads to improved efficacy with low computational cost.</li><li>The strategy is supported by theoretical justifications and empirical evidence.</li></ul>

Sharpness-Aware Parameter Selection for Machine Unlearning

Discover more