Associate Professor Nobuyuki Yoshioka of the International Center for Elementary Particle Physics at The University of Tokyo; graduate student Kento Tsubouchi of the Graduate School of Engineering at The University of Tokyo; and graduate student Haipeng Xie and Professor Ying Li of the Graduate School of the China Academy of Engineering Physics have proposed Argument Reweighting (AR), a candidate-solution reweighting method that ‘shakes up’ the decoding results of quantum error correction to test their reliability. In AR, the algorithm’s weights are changed so that the first correction candidate becomes less likely, and the same data are decoded again. If the decoding result does not change, it is accepted as a valid result; if it changes, it is discarded—an operation known as post-selection. Because the proposed method is extremely simple, it can be applied to a wide range of quantum error-correcting codes and decoding algorithms. It is also highly effective: for a family of codes known as BB codes, the researchers showed that discarding only about one result in 70,000 can reduce the logical error rate to one-sixtieth or less. Given that quantum error correction is used in a broad range of quantum-information technologies, including quantum communication, quantum sensing, and quantum cryptography, the findings are expected to have wide-ranging impact.
Conceptual illustration of the proposed method. The reliability of a computational result is quantitatively evaluated from the stability of the outputs obtained by running quantum-error-correction decoding several times.
Papers
Journal: Physical Review Letters
Title: Simple, Efficient, and Generic Post-Selection Decoding for qLDPC Codes
Authors: Haipeng Xie, Nobuyuki Yoshioka, Kento Tsubouchi, and Ying Li
DOI: 10.1103/sv11-f78l