Constructing an approximate logical Markovian model of consecutive QEC cycles of a stabilizer code

Abstract

As quantum error correction (QEC) experiments continue to make rapid progress, there is increased interest in designing experiments with guarantees of logical performance. At present, one difficulty is the lack of a clear connection between logical performance and the low-level error models. In this work, we take an important step toward addressing this issue by proving that consecutive QEC cycles of a stabilizer code with Pauli stochastic noise and with a single-cycle infidelity $ε_1 \leq 1/64$ admit an approximate logical Markovian model, meaning that consecutive noisy QEC cycles can be modeled by a memoryless error process acting only on the logical subsystem. The approximate logical Markovian model can be computed from the low-level error model, and the deviations from the true behavior are exponentially suppressed in the number of QEC cycles. Consequently, we expect that the approximate logical Markovian model will be both a useful tool for logical characterization and an aid for designing stabilizer-code implementations with guarantees of logical performance.