Abstract:We discuss the evaluation of the integrals for intermediate-order diagrams in the self-consistent strong-coupling expansion on the Keldysh contour using Tensor Cross Interpolation (TCI). TCI is used to factorize the nested parts of the integrand, allowing the integral to be computed as a recursion of convolution integrals, which are efficiently evaluated using the Fast Fourier Transform. The evaluation of diagrams where all vertices lie on one branch of the Keldysh contour resembles the structure of the imaginary-time formalism. For diagrams with time arguments on both contour branches, we find that it can be advantageous to parametrize the integrals in terms of physical time arguments with an additional sum over Keldysh indices. We benchmark the solution in relevant test cases, including the single impurity Anderson model and an exactly solvable electron-boson model. While the bond dimension increases with diagram order, the TCI-based integration efficiently handles low-order diagrams, making it a promising approach to go beyond the non-crossing approximation in steady-state non-equilibrium dynamical mean-field theory simulations.

What problem does this paper attempt to address?

This paper attempts to solve the problem of multi - dimensional integral calculation in the quantum impurity model under non - equilibrium steady - state, especially the integral evaluation of intermediate - order graphs in the strong - coupling expansion. Specifically, the paper focuses on how to use the Tensor Cross Interpolation (TCI) method to calculate these integrals efficiently. The following are the main objectives and problems solved in the paper: 1. **Efficient Calculation of Multi - Dimensional Integrals**: - The paper proposes to use the TCI method to decompose and calculate multi - dimensional integrals in the strong - coupling expansion. TCI realizes efficient numerical calculation by representing the integral function as a Matrix Product State (MPS). - Especially on the Keldysh contour, TCI can transform complex multi - dimensional integrals into recursive convolution integrals and use the Fast Fourier Transform (FFT) for efficient calculation. 2. **Handling Different Types of Graphs**: - For graphs with all vertices on the same branch of the Keldysh contour, whose structure is similar to that of graphs in the imaginary - time formalism, TCI can be directly applied for calculation. - For graphs with time parameters on two branches of the Keldysh contour, the paper proposes a new parameterization method, that is, adding the sum of Keldysh indices on the basis of physical time parameterization to improve the calculation efficiency. 3. **Benchmark Testing and Verification**: - The paper benchmarks the TCI method through several specific test cases, including the single - impurity Anderson model and an exactly solvable electron - boson model. - The test results show that although the bond dimension of MPS will increase as the graph order increases, the TCI method can still efficiently handle low - order graphs, providing a method beyond the non - crossing approximation (NCA) for non - equilibrium steady - state dynamical mean - field theory (DMFT) simulations. 4. **Parameterization Selection**: - The paper explores the influence of different parameterization methods on the calculation efficiency. For example, for the self - energy Σ< less than a part, the traditional loop parameterization may require a higher matrix dimension, while the Keldysh parameterization based on the physical time difference can more effectively handle integrals in a large time range. In conclusion, this paper aims to solve the bottleneck problem of multi - dimensional integral calculation in the non - equilibrium steady - state quantum impurity model by introducing and optimizing the TCI method, and provide an efficient and accurate solution for non - equilibrium DMFT simulations.

Solving quantum impurity models in the non-equilibrium steady state with tensor trains

A Tensor Train Continuous Time Solver for Quantum Impurity Models

Real-time evolution of Anderson impurity models via tensor network influence functionals

Strong coupling impurity solver based on quantics tensor cross interpolation

Solving quantum impurity problems on the L-shaped Kadanoff-Baym contour

Two-particle calculations with quantics tensor trains -- solving the parquet equations

Real-time Impurity Solver Using Grassmann Time-Evolving Matrix Product Operators

Nonequilibrium diagrammatic many-body simulations with quantics tensor trains

Low-rank quantics tensor train representations of Feynman diagrams for multiorbital electron-phonon models

Grassmann Time-Evolving Matrix Product Operators for Equilibrium Quantum Impurity Problems

Infinite Grassmann time-evolving matrix product operators for non-equilibrium quantum impurity problems

Anderson impurity solver integrating tensor network methods with quantum computing

Performance analysis of continuous-time solvers for quantum impurity models

Strong coupling quantum impurity solver on the real and imaginary axis

Continuous-time solver for quantum impurity models

CT-X: an Efficient Continuous-Time Quantum Monte Carlo Impurity Solver in the Kondo Regime.

Tensor network influence functionals in the continuous-time limit: connections to quantum embedding, bath discretization, and higher-order time propagation

Grassmann Time-Evolving Matrix Product Operators for Quantum Impurity Models

Grassmann time-evolving matrix product operators: An efficient numerical approach for fermionic path integral simulations

Equilibrium Quantum Impurity Problems via Matrix Product State Encoding of the Retarded Action