Membership Inference Attacks

Overfitting Vulnerability

Increased susceptibility to membership inference attacks when the model overfits its training data, producing distinct predictions for seen and unseen examples.

← Tillbaka