EFFICIENT QUANTUM STATE RECONSTRUCTION USING UNSUPERVISED LEARNING FOR QUANTUM CIRCUIT CUTTING
Keywords: Unsupervised learning, state tomography, quantum computing
TL;DR: Efficient quantum state tomography using RBM for quantum circuit cutting
Abstract: Current quantum computer (QC) fabrication encounters challenges when attempting to scale up the number of qubits. These challenges include errors, physical limitations, interference, and various other factors. As a remedy, quantum circuit cutting holds the promise for studying large quantum systems with the limited qubit capacity of quantum computers today. With quantum circuit cutting, the output of a large quantum circuit could be obtained through classical post-processing of fragmented circuit outputs acquired through different measurement and preparation bases. However, such reconstruction process results in exponential quantum measurement cost with the increase in the number of circuit cuts. In this paper, we demonstrate efficient state reconstruction using a Restricted Boltzmann Machine (RBM) with polynomial resource scaling. We explore the benefits of unsupervised learning for simulating extensive quantum systems, exemplified by the reconstruction of highly entangled multi-qubit Greenberger–Horne–Zeilinger (GHZ) states from fragmented circuits. Our experiments illustrate that fragmented GHZ circuits, at the state-of-the-art scale of up to $18$ qubits, can be reconstructed with near-perfect fidelity using only $100$ sample measurements compared to $4^{18}$ sample measurements needed otherwise.
Primary Area: applications to physical sciences (physics, chemistry, biology, etc.)
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 6541
Loading