# Decoding probabilistic syndrome measurement and the role of entropy

## TL;DR

This paper investigates the performance of the toric code under probabilistic stabiliser measurements, demonstrating that high error thresholds are achievable with modified decoding methods, and explores the impact of entropy on decoding efficiency.

## Contribution

It introduces a modified decoding approach for probabilistic stabiliser measurements and analyzes the role of entropy in improving decoder performance.

## Key findings

- High error threshold of 1.69% with probabilistic measurements using edge-contraction decoding
- Deterministic measurements achieve a higher threshold of 2.93%
- Entropy factors have negligible advantage in fully continuous measurement scenarios

## Abstract

In realistic stabiliser-based quantum error correction there are many ways in which real physical systems deviate from simple toy models of error. Stabiliser measurements may not always be deterministic or may suffer from erasure errors, such that they do not supply syndrome outcomes required for error correction. In this paper, we study the performance of the toric code under a model of probabilistic stabiliser measurement. We find that, even under a completely continuous model of syndrome extraction, the threshold can be maintained at reasonably high values of $1.69\%$ by suitably modifying the decoder using the edge-contraction method of Stace and Barrett (Physical Review A 81, 022317 (2010)), compared to a value of $2.93\%$ for deterministic stabiliser measurements. Finally, we study the role of entropic factors which account for degenerate error configurations for improving on the performance of the decoder. We find that in the limit of completely continuous stabiliser measurement any advantage further provided by these factors becomes negligible in contrast to the case of deterministic measurements.