Unsupervised Interpretable Learning of Phases From Many-Qubit Systems

TL;DR

This paper introduces an unsupervised machine learning approach to analyze quantum data from many-qubit systems, successfully identifying phases and order parameters without prior knowledge, aiding interpretability in quantum physics.

Contribution

It presents a novel unsupervised learning method that constructs phase diagrams and detects order parameters in many-qubit systems, enhancing interpretability without prior information.

Findings

Successfully constructs phase diagram of a cluster-state model

Detects string order parameters in quantum phases

Identifies explicit stabilizers in the toric code under magnetic fields

Abstract

Experimental progress in qubit manufacturing calls for the development of new theoretical tools to analyze quantum data. We show how an unsupervised machine-learning technique can be used to understand short-range entangled many-qubit systems using data of local measurements. The method successfully constructs the phase diagram of a cluster-state model and detects the respective order parameters of its phases, including string order parameters. For the toric code subject to external magnetic fields, the machine identifies the explicit forms of its two stabilizers. Prior information of the underlying Hamiltonian or the quantum states is not needed; instead, the machine outputs their characteristic observables. Our work opens the door for a first-principles application of hybrid algorithms that aim at strong interpretability without supervision.