Optimization schemes for unitary tensor-network circuit

TL;DR

This paper explores variational optimization techniques for unitary tensor-network circuits, comparing different structures through benchmarking on disordered Heisenberg models and tensor decompositions.

Contribution

It introduces a generalized variational optimization framework combining multi-scale entanglement renormalization and conjugate-gradient methods for tensor networks.

Findings

Effective optimization of tensor-network circuits demonstrated.

Benchmarking shows advantages of different network structures.

Method applicable to disordered quantum systems.

Abstract

We discuss the variational optimization of a unitary tensor-network circuit with different network structures. The ansatz is performed based on a generalization of well-developed multi-scale entanglement renormalization algorithm and also the conjugate-gradient method with an effective line search. We present the benchmarking calculations for different network structures by studying the Heisenberg model in a strongly disordered magnetic field and a tensor-network $QR$-decomposition.