From the 1 of 4 linked papers with an AI index.
4 papers
Generalization and Memorization in Rectified Flow
Mingxing Rao, Daniel Moyer
The paper examines how Rectified Flow generative models memorize training data by designing calibrated membership inference attacks, discovers peak privacy risk at the integration…
Score-based Membership Inference on Diffusion Models
Mingxing Rao, Bowen Qu, Daniel Moyer
Membership inference attacks (MIAs) against Diffusion Models (DMs) raise pressing privacy concerns by revealing whether a sample was part of the training set. While existing method…
Latent Diffusion Inversion Requires Understanding the Latent Space
Mingxing Rao, Bowen Qu, Daniel Moyer
The recovery of training data from generative models ("model inversion") has been extensively studied for diffusion models in the data domain as a memorization/overfitting phenomen…
Training Noise Token Pruning
Mingxing Rao, Bohan Jiang, Daniel Moyer
In the present work we present Training Noise Token (TNT) Pruning for vision transformers. Our method relaxes the discrete token dropping condition to continuous additive noise, pr…