QMoE: A Quantum Mixture of Experts Framework for Scalable Quantum Neural Networks

TL;DR

This paper introduces QMoE, a quantum mixture of experts framework that enhances the scalability and expressiveness of quantum neural networks by integrating multiple quantum experts with a learnable routing mechanism, demonstrating superior performance on classification tasks.

Contribution

It presents the first quantum mixture of experts architecture that combines multiple quantum circuits with a routing mechanism, improving scalability and performance in quantum machine learning.

Findings

QMoE outperforms standard quantum neural networks on classification tasks.

The framework demonstrates improved scalability and expressiveness.

QMoE offers a pathway for more interpretable quantum learning models.

Abstract

Quantum machine learning (QML) has emerged as a promising direction in the noisy intermediate-scale quantum (NISQ) era, offering computational and memory advantages by harnessing superposition and entanglement. However, QML models often face challenges in scalability and expressiveness due to hardware constraints. In this paper, we propose quantum mixture of experts (QMoE), a novel quantum architecture that integrates the mixture of experts (MoE) paradigm into the QML setting. QMoE comprises multiple parameterized quantum circuits serving as expert models, along with a learnable quantum routing mechanism that selects and aggregates specialized quantum experts per input. The empirical results from the proposed QMoE on quantum classification tasks demonstrate that it consistently outperforms standard quantum neural networks, highlighting its effectiveness in learning complex data patterns. Our work paves the way for scalable and interpretable quantum learning frameworks.