EM-MIAs: Enhancing Membership Inference Attacks in Large Language Models through Ensemble Modeling.

Zichen Song, Sitan Huang, Zhongfeng Kang

Browse the full ICASSP paper archive.