Powerful Training-Free Membership Inference Against Autoregressive Language Models

arXiv:2601.12104v2 Announce Type: replace-cross Abstract: Fine-tuned language models pose significant privacy risks, as they may memorize and expose sensitive information from their training data. Membership inference attacks (MIAs) provide a principled framework for auditing these risks, yet existing methods achieve limited detection rates, particularly at the low false-positive thresholds required for practical privacy auditing. We present EZ-MIA, a membership inference attack that exploits a