Hybrid Quantum Circuits

Andrew C. Potter, Romain Vasseur

Jérome Ricciardi, Sébastien Bardin, Christophe Chareton +1 more

Equivalence checking of hybrid quantum circuits is of primary importance, given that quantum circuit transformations are omnipresent along the quantum compiler chain. While some approaches exist for automating this task, most focus on the simple case of unitary circuits. At the same time, real quantum computing requires hybrid circuits equipped with measurement operators. Moreover, the few approaches targeting the hybrid case are limited to a restricted class of problems. We propose tackling the Quantum Hybrid Circuit Equivalence Checking problem through lifting unitary circuit verification using a transformation known as deferred measurement. We show that this approach alone significantly outperforms prior work, and that, with the addition of specific unitary-level techniques we call separation and projection, it can handle much larger classes of hybrid circuit equivalence problems. We have implemented and evaluated our method over standard circuit transformations such as teleportation, one-way measurement, or the IBM Qiskit compiler, demonstrating its promises. As a side finding, we have identified and reported several unexpected behaviours with the Qiskit compiler.

D. Vion, A. Aassime, Audrey Cottet +5 more

Yevhenii Trochun, Evgen Pavlov, Sergii Stirenko +1 more

This article describes hybrid convolutional neural network that uses one quantum circuit for image classification. The different configurations of the hybrid neural network with the quantum circuit are considered. Several different quantum circuits with different number of qubits are compared. These hybrid neural network configurations are evaluated on MNIST and MNIST Fashion datasets, which is radically different from MNIST dataset. Performance of hybrid neural network is compared for multiclass classification on MNIST and MNIST Fashion datasets for 4, 6, 8, 10 classes using quantum circuits with 2, 3, 4 qubits. The results of the experiments indicate the feasibility of using hybrid neural networks for multiclass classification.

DANIEL RIBAS TANDEITNIK, Thiago Guerreiro

François Chapeau‐Blondeau