Benchmarking a Tunable Quantum Neural Network on Trapped-Ion and Superconducting Hardware

TL;DR

This paper demonstrates a quantum neural network implemented on trapped-ion and superconducting hardware for image classification, showing quantum effects can enhance performance and reveal sensitivities to physical noise, paving the way for more complex quantum AI.

Contribution

It introduces a tunable quantum neural network on real hardware, analyzing quantum-classical interpolation effects and noise impacts on classification performance.

Findings

Quantum uncertainty improves classification at moderate levels.

Quantum networks detect certain images misclassified by classical networks.

Physical noise causes output fluctuations, especially in borderline cases.

Abstract

We implement a quantum generalization of a neural network on trapped-ion and IBM superconducting quantum computers to classify MNIST images, a common benchmark in computer vision. The network feedforward involves qubit rotations whose angles depend on the results of measurements in the previous layer. The network is trained via simulation, but inference is performed experimentally on quantum hardware. The classical-to-quantum correspondence is controlled by an interpolation parameter, $a$, which is zero in the classical limit. Increasing $a$ introduces quantum uncertainty into the measurements, which is shown to improve network performance at moderate values of the interpolation parameter. We then focus on particular images that fail to be classified by a classical neural network but are detected correctly in the quantum network. For such borderline cases, we observe strong deviations from the simulated behavior. We attribute this to physical noise, which causes the output to fluctuate between nearby minima of the classification energy landscape. Such strong sensitivity to physical noise is absent for clear images. We further benchmark physical noise by inserting additional single-qubit and two-qubit gate pairs into the neural network circuits. Our work provides a springboard toward more complex quantum neural networks on current devices: while the approach is rooted in standard classical machine learning, scaling up such networks may prove classically non-simulable and could offer a route to near-term quantum advantage.