#machine learning
May 2026
Fast Adversarial Attacks with Gradient Prediction
A family of attacks is introduced that eliminates the backward pass by predicting the input gradient from forward-pass hidden states via a lightweight linear regression, and exact affine conditional gradient means are derived.
Kamil Ciosek, Aleksandr V. Petrov, Nicolò Felicioni et al.
· arXiv.org · 1 citation