Quantum Diffusion Models

TL;DR

This paper introduces a quantum generative diffusion model that uses parameterized quantum circuits to generate quantum states, with implementations tested on real quantum hardware and various model variants explored.

Contribution

It presents a novel quantum diffusion framework replacing neural networks with quantum circuits, including full, latent, and conditioned versions, and demonstrates practical implementation on NISQ devices.

Findings

Quantum diffusion models can generate quantum states effectively.

The models perform well according to quantitative and qualitative metrics.

A simplified version was successfully implemented on real quantum hardware.

Abstract

We propose a quantum version of a generative diffusion model. In this algorithm, artificial neural networks are replaced with parameterized quantum circuits, in order to directly generate quantum states. We present both a full quantum and a latent quantum version of the algorithm; we also present a conditioned version of these models. The models' performances have been evaluated using quantitative metrics complemented by qualitative assessments. An implementation of a simplified version of the algorithm has been executed on real NISQ quantum hardware.