Metaplectic Anyons, Majorana Zero Modes, and their Computational Power

TL;DR

This paper introduces a new class of anyon models combining Ising anyons with $SO(m)_2$ Chern-Simons theory, analyzing their quasiparticles, braiding properties, and computational complexity, revealing non-universality but computational hardness of certain invariants.

Contribution

It presents a novel anyon model generalizing Majorana modes, with detailed analysis of its quasiparticles, braiding, and computational complexity, including classical simulability and #P-hardness results.

Findings

Braid group image is finite and non-universal for quantum computation.

Fundamental quasiparticle braiding can be classically simulated.

Certain composite quasiparticle braiding is #P-hard.

Abstract

We introduce and study a class of anyon models that are a natural generalization of Ising anyons and Majorana fermion zero modes. These models combine an Ising anyon sector with a sector associated with $SO(m)_2$ Chern-Simons theory. We show how they can arise in a simple scenario for electron fractionalization and give a complete account of their quasiparticles types, fusion rules, and braiding. We show that the image of the braid group is finite for a collection of $2n$ fundamental quasiparticles and is a proper subgroup of the metaplectic representation of $Sp(2n-2,\mathbb{F}_m)\ltimes H(2n-2,\mathbb{F}_m)$, where $Sp(2n-2,\mathbb{F}_m)$ is the symplectic group over the finite field $\mathbb{F}_m$ and $H(2n-2,\mathbb{F}_m)$ is the extra special group (also called the $(2n-1)$-dimensional Heisenberg group) over $\mathbb{F}_m$. Moreover, the braiding of fundamental quasiparticles can be efficiently simulated classically. However, computing the result of braiding a certain type of composite quasiparticle is $# P$-hard, although it is not universal for quantum computation because it has a finite braid group image. This a rare example of a topological phase that is not universal for quantum computation through braiding but nevertheless has $# P$-hard link invariants. We argue that our models are closely related to recent analyses finding non-Abelian anyonic properties for defects in quantum Hall systems, generalizing Majorana zero modes in quasi-1D systems.