Scalable measurement error mitigation via iterative bayesian unfolding

doi:10.1103/PhysRevResearch.6.013187

Scalable measurement error mitigation via iterative bayesian unfolding

2024-02-21

3

OA

AI

DOI：10.1103/PhysRevResearch.6.013187

Original

Original request for help

Save

Abstract

En 中文

Measurement errors are a significant obstacle to achieving scalable quantum computation. To counteract systematic readout errors, researchers have developed postprocessing techniques known as measurement error mitigation methods. However, these methods face a tradeoff between scalability and returning nonnegative probabilities. In this paper, we present a solution to overcome this challenge. Our approach focuses on iterative Bayesian unfolding, a standard mitigation technique used in high-energy physics experiments, and implements it in a scalable way. We demonstrate our method on experimental Greenberger-Horne-Zeilinger state preparation on up to 127 qubits and on the Bernstein-Vazirani algorithm on up to 26 qubits. Compared to state-of-the-art methods (such as M3), our implementation guarantees valid probability distributions, returns comparable or better -mitigated results, and does so without a noticeable time and memory overhead.

Scalable measurement error mitigation via iterative bayesian unfolding

Abstract

Journal

Organization