Universal Quantum Tomography With Deep Neural Networks
Nhan T. Luu, Thang C. Truong, Duong T. Luu
TL;DR
This work addresses the challenge of reconstructing both pure and mixed quantum states via quantum state tomography using deep neural networks. It introduces two architectures, RFB-Net and MS-NN, designed to operate on 32x32 density-matrix inputs and handle diverse state classes, including bosonic codes such as GKP and cat states. Through a combination of convolutional feature extraction, multitask outputs, and a Cholesky-based density-operator construction, the methods achieve high fidelity reconstructions, outperforming a prior QST-CGAN baseline especially in mixed-state scenarios, and demonstrate robustness under realistic noise models. The results suggest a scalable, data-driven pathway to practical quantum tomography for intermediate-scale quantum systems with potential applications in quantum information processing and hardware verification.
Abstract
Quantum state tomography is a crucial technique for characterizing the state of a quantum system, which is essential for many applications in quantum technologies. In recent years, there has been growing interest in leveraging neural networks to enhance the efficiency and accuracy of quantum state tomography. Still, many of them did not include mixed quantum state, since pure states are arguably less common in practical situations. In this research paper, we present two neural networks based approach for both pure and mixed quantum state tomography: Restricted Feature Based Neural Network and Mixed States Conditional Generative Adversarial Network, evaluate its effectiveness in comparison to existing neural based methods. We demonstrate that our proposed methods can achieve state-of-the-art results in reconstructing mixed quantum states from experimental data. Our work highlights the potential of neural networks in revolutionizing quantum state tomography and facilitating the development of quantum technologies.