Title:Accelerated Inchworm Method with Tensor-Train Bath Influence Functional

Abstract:We propose an efficient tensor-train-based algorithm for simulating open quantum systems with the inchworm method, where the reduced dynamics of the open quantum system is expressed as a perturbative series of high-dimensional integrals. Instead of evaluating the integrals with Monte Carlo methods, we approximate the costly bath influence functional (BIF) in the integrand as a tensor train, allowing accurate deterministic numerical quadrature schemes implemented in an iterative manner. Thanks to the low-rank structure of the tensor train, our proposed method has a complexity that scales linearly with the number of dimensions. Our method couples seamlessly with the tensor transfer method, allowing long-time simulations of the dynamics.