Towards a scalable discrete quantum generative adversarial neural network

TL;DR

This paper presents a novel fully quantum generative adversarial network designed for binary data, combining innovative features to enhance expressivity and demonstrate potential for data reproduction and generalization.

Contribution

It introduces a fully quantum GAN architecture with integrated features like noise reuploading and auxiliary qubits, not previously combined in quantum models.

Findings

Successfully reproduces synthetic and Ising model data

Demonstrates potential for generalization from training data

Shows the combined quantum features enhance model expressivity

Abstract

We introduce a fully quantum generative adversarial network intended for use with binary data. The architecture incorporates several features found in other classical and quantum machine learning models, which up to this point had not been used in conjunction. In particular, we incorporate noise reuploading in the generator, auxiliary qubits in the discriminator to enhance expressivity, and a direct connection between the generator and discriminator circuits, obviating the need to access the generator's probability distribution. We show that, as separate components, the generator and discriminator perform as desired. We empirically demonstrate the expressive power of our model on both synthetic data as well as low energy states of an Ising model. Our demonstrations suggest that the model is not only capable of reproducing discrete training data, but also of potentially generalizing from it.