Skip to yearly menu bar Skip to main content


Poster

Effect of Ambient-Intrinsic Dimension Gap on Adversarial Vulnerability

Rajdeep Haldar · Yue Xing · Qifan Song

MR1 & MR2 - Number 80
[ ]
Thu 2 May 8 a.m. PDT — 8:30 a.m. PDT

Abstract: The existence of adversarial attacks on machine learning models imperceptible to a human is still quite a mystery from a theoretical perspective. In this work, we introduce two notions of adversarial attacks: natural or on-manifold attacks, which are perceptible by a human/oracle, and unnatural or off-manifold attacks, which are not. We argue that the existence of the off-manifold attacks is a natural consequence of the dimension gap between the intrinsic and ambient dimensions of the data. For 2-layer ReLU networks, we prove that even though the dimension gap does not affect generalization performance on samples drawn from the observed data space, it makes the clean-trained model more vulnerable to adversarial perturbations in the off-manifold direction of the data space. Our main results provide an explicit relationship between the $\ell_2,\ell_{\infty}$ attack strength of the on/off-manifold attack and the dimension gap.

Live content is unavailable. Log in and register to view live content