Quantum Brain
← Back to papers

Universality Classes for Purification in Nonunitary Quantum Processes

Andrea De Luca, Chunxiao Liu, Adam Nahum, Tianci Zhou·December 29, 2023·DOI: 10.1103/wlj6-mkk4
Physics

AI Breakdown

Get a structured breakdown of this paper — what it's about, the core idea, and key takeaways for the field.

Abstract

We consider the universal aspects of two problems: (i) the singular value structure of a product M t = m t m t − 1 … m 1 of many large independent random matrices and (ii) the slow purification of a large number of qubits by repeated quantum measurements. The time-evolution operator in the latter case is again a product of matrices m i , representing time steps in the evolution, but the m i are now nontrivially correlated as a result of Born’s rule. Both processes are associated with the decay of natural measures of entropy as a function of time or of the number of matrices in the product. We argue that, for a broad class of models, each process is described by universal scaling forms for purification and that (i) and (ii) represent distinct “universality classes” with distinct scaling functions. Using the replica trick, these universality classes correspond to effective one-dimensional statistical mechanics models for a gas of “kinks,” representing domain walls between elements of the permutation group. This is an instructive low-dimensional limit of the effective statistical mechanics models for random circuits and tensor networks. These results apply to longtime purification in spatially local monitored circuit models on the entangled side of the measurement phase transition.

Related Research

Quantum Intelligence

Ask about quantum research, companies, or market developments.