Categorized Privacy Neuron Editing: Addressing Heterogeneity for Privacy Leakage Mitigation in Language Models.

Xiaole Yang, Peng Wu

Browse the full ADMA paper archive.