Universal Adversarial Examples and Perturbations for Quantum Classifiers

TL;DR

This paper investigates the existence and properties of universal adversarial examples in quantum classifiers, demonstrating their ability to fool multiple models and analyzing the perturbation strength needed for such attacks.

Contribution

It introduces the concept of universal adversarial examples for quantum classifiers and provides theoretical bounds on the perturbation strength required for successful attacks.

Findings

Universal adversarial examples can fool multiple quantum classifiers.

A logarithmic increase in perturbation strength relative to the number of classifiers is sufficient.

Existence of universal adversarial perturbations for individual quantum classifiers.

Abstract

Quantum machine learning explores the interplay between machine learning and quantum physics, which may lead to unprecedented perspectives for both fields. In fact, recent works have shown strong evidences that quantum computers could outperform classical computers in solving certain notable machine learning tasks. Yet, quantum learning systems may also suffer from the vulnerability problem: adding a tiny carefully-crafted perturbation to the legitimate input data would cause the systems to make incorrect predictions at a notably high confidence level. In this paper, we study the universality of adversarial examples and perturbations for quantum classifiers. Through concrete examples involving classifications of real-life images and quantum phases of matter, we show that there exist universal adversarial examples that can fool a set of different quantum classifiers. We prove that for a set of $k$ classifiers with each receiving input data of $n$ qubits, an $O(\frac{\ln k} {2^n})$ increase of the perturbation strength is enough to ensure a moderate universal adversarial risk. In addition, for a given quantum classifier we show that there exist universal adversarial perturbations, which can be added to different legitimate samples and make them to be adversarial examples for the classifier. Our results reveal the universality perspective of adversarial attacks for quantum machine learning systems, which would be crucial for practical applications of both near-term and future quantum technologies in solving machine learning problems.