Trustworthy Quantum Artificial Intelligence

In the Trustworthy Quantum Artificial Intelligence project in VeriQ, we explore new methods and develop a series of tools to verify Quantum Artificial Intelligence (QAI) models such as quantum deep neural networks and quantum convolutional neural networks. On the basis of our defined formal frameworks, these tools can not only automatically check the trustworthy properties (e.g., robustness and fairness) of QAI models, but also provide counterexamples indicating how the models are not satisfied with the properties.

Robustness Verifier

A python toolbox for robustness verification of quantum classifiers.

See Github Repo

[1] Ji Guan, Wang Fang, and Mingsheng Ying. Robustness verification of quantum classifiers. In Computer Aided Verification - 33rd International Conference, CAV 2021, volume 12759 of Lecture Notes in Computer Science, pages 151–174. Springer, 2021. [ DOI | http ]

Fairness Verifier

A python toolbox for fairness verification of quantum decision models by computing the Lipschitz constant of the models.

See Github Repo