Variational optimization of tensor-network states with the honeycomb-lattice corner transfer matrix

TL;DR

This paper introduces a variational optimization method for tensor-network states on the honeycomb lattice, utilizing automatic differentiation of corner transfer matrix renormalization group, achieving accurate results for quantum spin models.

Contribution

The paper presents a novel variational optimization technique for tensor-network states on the honeycomb lattice using automatic differentiation, enabling precise calculations for complex quantum models.

Findings

Accurately computes physical observables for Heisenberg and Kitaev models.

Demonstrates the effectiveness of the method on honeycomb lattice models.

Provides a foundation for future extensions of tensor-network algorithms.

Abstract

We develop a method of variational optimization of the infinite projected entangled pair states on the honeycomb lattice. The method is based on the automatic differentiation of the honeycomb-lattice corner transfer matrix renormalization group. We apply the approach to the antiferromagnetic Heisenberg spin-1/2 and ferromagnetic Kitaev models on the honeycomb lattice. The developed formalism gives quantitatively accurate results for the main physical observables and has a necessary potential for further extensions.