Regularized scheme of time evolution tensor network algorithms

TL;DR

This paper introduces a regularized tensor network algorithm for simulating quantum lattice system time evolution, offering improved stability and efficiency over traditional methods, especially for complex spin models.

Contribution

It develops a novel regularized factorization scheme that surpasses Trotter decomposition, enabling stable and cost-effective ground state calculations for spin lattice systems.

Findings

Stable convergence immune to bias in simple update methods

Rapid tensor network contraction with lower computational cost

Effective simulation of Heisenberg and Kitaev spin systems

Abstract

Regularized factorization is proposed to simulate time evolution for quantum lattice systems. Transcending the Trotter decomposition, the resulting compact structure of the propagator indicates a high-order Baker-Campbell-Hausdorff series. Regularized scheme of tensor network algorithms is then developed to determine the ground state energy for spin lattice systems with Heisenberg or Kitaev-type interactions. Benchmark calculations reveal two distinct merits of the regularized algorithm: it has stable convergence, immune to the bias even in applying the simple update method to the Kitaev spin liquid; contraction of the produced tensor network can converge rapidly with much lower computing cost, relaxing the bottleneck to calculate the physical expectation value.