QRENN_Codes

The experimental codes and programs for QRENN project

Table of Contents

1. Overview
2. Repository Structure
3. Getting Started
4. Configuration
5. Evaluation
6. Contact

---

Overview

This is a Github repository for the academic research of 'Quantum Recurrent Embedding Neural Networks (QRENN)' describing a new design of quantum neural network design.

• Context:
  Focusing on the quantum supervised learning problems in the domain of quantum neural networks (QNNs) and qnantum machine learning (QML).
• Key Findings:
  1. We design a new QNN model in the framework of quantum data reuploading, based on the advanced quantum algorithm design, e.g., quantum singular value transformation (QSVT).
  2. We theoretically showcase that the model can avoid barren plateaus issue in quantum supervised learning tasks.
  3. We numerically showcase that QRENN model can be applied in Hamiltonian classification and symmetry-protected topological (SPT) detection tasks.
• Paper:
  arXiv version: Quantum Recurrent Embedding Neural Networks.

Repository Structure

QRENN_Codes

• application
  • unitary/Hamiltonian classification notebooks/python scripts
  • SPT detection data generation
  • SPT detection training
• trainability
  • gradient sampling python scripts for diagonal/involutory/pauli datasets.
• README.md # This file

Getting Started

Prerequisites

• Language/runtime (e.g., Python 3.8 or higher version, CUDA 12.1, etc.)
• QuAIRKit python QML toolkits (version 0.4.0 or higher)
• System requirements (CPU/GPU, RAM, OS)

Installation

Please visit web: QuAIR-Platform

Citation

@article{jing2025quantum,
  title={Quantum Recurrent Embedding Neural Network},
  author={Jing, Mingrui and Huang, Erdong and Shi, Xiao and Zhang, Shengyu and Wang, Xin},
  journal={arXiv preprint arXiv:2506.13185},
  year={2025}
}

Contact

See emails in the main paper