Stability and Loop Models from Decohering Non-Abelian Topological Order

TL;DR

This paper develops statistical mechanical models to understand decohering non-Abelian topological order, revealing stability properties and phase behavior relevant for quantum error correction in topological quantum memories.

Contribution

It introduces a framework linking decohering non-Abelian topological order to loop models with quantum dimension-dependent weights, providing exact and numerical analysis of stability and phase transitions.

Findings

Loop models describe decohered states with quantum dimension N

Stability increases with larger quantum dimensions of anyons

Critical phases may occur for smaller quantum dimensions

Abstract

Decohering topological order (TO) is central to the many-body physics of open quantum matter and decoding transitions. We identify relevant statistical mechanical models for decohering non-Abelian TO, which have been crucial for understanding the error threshold of Abelian stabilizer codes. The decohered density matrix can be described by loop models, whose topological loop weight $N$ is given by the quantum dimension of the decohering anyon -- reducing to the Ising model if $N=1$. In particular, the R\'enyi-$n$ moments of the decohered state correspond to $n$ coupled O$(N)$ loop models, and we exactly diagonalize the density matrix at maximal error rate. This allows us to relate the fidelity between two logically distinct ground states to properties of random O$(N)$ loop and spin models. Utilizing the literature on loop models, we find a remarkable stability to quantum channels which proliferate non-Abelian anyons with large quantum dimension, with the possibility of critical phases for smaller dimensions. We confirm our framework with exact results for Kitaev quantum double models, and with numerical simulations for the non-Abelian phase of the Kitaev honeycomb model. The latter is an example of a non-fixed-point wavefunction with non-bosonic and non-integral anyon dimensions. Our work opens up the possibility of non-Abelian TO being robust against maximally proliferating certain anyons, which can inform error-correction studies of these topological memories.