-
Learning tensor networks with tensor cross interpolation: new algorithms and libraries
Authors:
Yuriel Núñez Fernández,
Marc K. Ritter,
Matthieu Jeannin,
Jheng-Wei Li,
Thomas Kloss,
Thibaud Louvet,
Satoshi Terasaki,
Olivier Parcollet,
Jan von Delft,
Hiroshi Shinaoka,
Xavier Waintal
Abstract:
The tensor cross interpolation (TCI) algorithm is a rank-revealing algorithm for decomposing low-rank, high-dimensional tensors into tensor trains/matrix product states (MPS). TCI learns a compact MPS representation of the entire object from a tiny training data set. Once obtained, the large existing MPS toolbox provides exponentially fast algorithms for performing a large set of operations. We di…
▽ More
The tensor cross interpolation (TCI) algorithm is a rank-revealing algorithm for decomposing low-rank, high-dimensional tensors into tensor trains/matrix product states (MPS). TCI learns a compact MPS representation of the entire object from a tiny training data set. Once obtained, the large existing MPS toolbox provides exponentially fast algorithms for performing a large set of operations. We discuss several improvements and variants of TCI. In particular, we show that replacing the cross interpolation by the partially rank-revealing LU decomposition yields a more stable and more flexible algorithm than the original algorithm. We also present two open source libraries, xfac in Python/C++ and TensorCrossInterpolation.jl in Julia, that implement these improved algorithms, and illustrate them on several applications. These include sign-problem-free integration in large dimension, the superhigh-resolution quantics representation of functions, the solution of partial differential equations, the superfast Fourier transform, the computation of partition functions, and the construction of matrix product operators.
△ Less
Submitted 4 July, 2024; v1 submitted 2 July, 2024;
originally announced July 2024.
-
Matrix product states and first quantization
Authors:
Jheng-Wei Li,
Xavier Waintal
Abstract:
Common wisdom says that the entanglement of fermionic systems can be low in the second quantization formalism but is extremely large in the first quantization. Hence Matrix Product State (MPS) methods based on moderate entanglement have been overwhelmingly formulated in second quantization. Here we introduce a first-quantized MPS approach to simulate quantum many-body systems. We show that by refo…
▽ More
Common wisdom says that the entanglement of fermionic systems can be low in the second quantization formalism but is extremely large in the first quantization. Hence Matrix Product State (MPS) methods based on moderate entanglement have been overwhelmingly formulated in second quantization. Here we introduce a first-quantized MPS approach to simulate quantum many-body systems. We show that by reformulating the way the fermionic anti-symmetry is handled, we arrive at MPS with a level of entanglement comparable to the usual one found in second quantization. We demonstrate our scheme on the one-dimensional $t$-$V$ model (spinless fermions with nearest neighbour density-density interaction) for both ground state and time evolution. For time evolution, we find that the entanglement entropy in first quantization is significantly smaller than in its second quantization counterpart.
△ Less
Submitted 10 April, 2024;
originally announced April 2024.
-
Tensorized orbitals for computational chemistry
Authors:
Nicolas Jolly,
Yuriel Núñez Fernández,
Xavier Waintal
Abstract:
Choosing a basis set is the first step of a quantum chemistry calculation and it sets its maximum accuracy. This choice of orbitals is limited by strong technical constraints as one must be able to compute a large number of high dimensional integrals from these orbitals. Here we use tensor network techniques to construct representations of orbitals that essentially lift these technical constraints…
▽ More
Choosing a basis set is the first step of a quantum chemistry calculation and it sets its maximum accuracy. This choice of orbitals is limited by strong technical constraints as one must be able to compute a large number of high dimensional integrals from these orbitals. Here we use tensor network techniques to construct representations of orbitals that essentially lift these technical constraints. We show that a large class of orbitals can be put into ``tensorized'' form including the Gaussian orbitals, the exact hydrogenoid atomic orbitals and orbitals for chemical bonds. Our method provides a path for building more accurate and more compact basis sets beyond what has been accessible with previous technology.
△ Less
Submitted 7 August, 2023;
originally announced August 2023.
-
Quantics Tensor Cross Interpolation for High-Resolution, Parsimonious Representations of Multivariate Functions in Physics and Beyond
Authors:
Marc K. Ritter,
Yuriel Núñez Fernández,
Markus Wallerberger,
Jan von Delft,
Hiroshi Shinaoka,
Xavier Waintal
Abstract:
Multivariate functions of continuous variables arise in countless branches of science. Numerical computations with such functions typically involve a compromise between two contrary desiderata: accurate resolution of the functional dependence, versus parsimonious memory usage. Recently, two promising strategies have emerged for satisfying both requirements: (i) The quantics representation, which e…
▽ More
Multivariate functions of continuous variables arise in countless branches of science. Numerical computations with such functions typically involve a compromise between two contrary desiderata: accurate resolution of the functional dependence, versus parsimonious memory usage. Recently, two promising strategies have emerged for satisfying both requirements: (i) The quantics representation, which expresses functions as multi-index tensors, with each index representing one bit of a binary encoding of one of the variables; and (ii) tensor cross interpolation (TCI), which, if applicable, yields parsimonious interpolations for multi-index tensors. Here, we present a strategy, quantics TCI (QTCI), which combines the advantages of both schemes. We illustrate its potential with an application from condensed matter physics: the computation of Brillouin zone integrals.
△ Less
Submitted 25 March, 2024; v1 submitted 21 March, 2023;
originally announced March 2023.
-
The self-consistent quantum-electrostatic problem in strongly non-linear regime
Authors:
P. Armagnat,
A. Lacerda-Santos,
B. Rossignol,
C. Groth,
X. Waintal
Abstract:
The self-consistent quantum-electrostatic (also known as Poisson-Schrödinger) problem is notoriously difficult in situations where the density of states varies rapidly with energy. At low temperatures, these fluctuations make the problem highly non-linear which renders iterative schemes deeply unstable. We present a stable algorithm that provides a solution to this problem with controlled accuracy…
▽ More
The self-consistent quantum-electrostatic (also known as Poisson-Schrödinger) problem is notoriously difficult in situations where the density of states varies rapidly with energy. At low temperatures, these fluctuations make the problem highly non-linear which renders iterative schemes deeply unstable. We present a stable algorithm that provides a solution to this problem with controlled accuracy. The technique is intrinsically convergent including in highly non-linear regimes. We illustrate our approach with (i) a calculation of the compressible and incompressible stripes in the integer quantum Hall regime and (ii) a calculation of the differential conductance of a quantum point contact geometry. Our technique provides a viable route for the predictive modeling of the transport properties of quantum nanoelectronics devices.
△ Less
Submitted 30 August, 2019; v1 submitted 3 May, 2019;
originally announced May 2019.
-
Numerical cookbook for electronic quantum transport at finite frequency
Authors:
Oleksii Shevtsov,
Xavier Waintal
Abstract:
Building on the many existing algorithms for calculating the DC transport properties of quantum tight-binding models, we develop a systematic approach that expresses finite frequency observables in terms of the stationary Green's function of the system, i.e. the natural output of most DC numerical codes. Our framework allows to extend the simulations capabilities of existing codes to a large class…
▽ More
Building on the many existing algorithms for calculating the DC transport properties of quantum tight-binding models, we develop a systematic approach that expresses finite frequency observables in terms of the stationary Green's function of the system, i.e. the natural output of most DC numerical codes. Our framework allows to extend the simulations capabilities of existing codes to a large class of observables including, for instance, AC conductance, quantum capacitance, quantum pumping, spin pumping or photo-assisted shot noise. The theory is developed within the framework of Keldysh formalism and we provide explicit links with the alternative (and equivalent) scattering approach. We illustrate the formalism with a study of the AC conductance in a quantum point contact and an electronic Mach-Zehnder interferometer in the quantum Hall regime.
△ Less
Submitted 12 November, 2012;
originally announced November 2012.
-
KNIT : An open source code for quantum transport in multi-terminal systems
Authors:
Irina Rychkova,
Valentin Rychkov,
Kyryl Kazymyrenko,
Simone Borlenghi,
Xavier Waintal
Abstract:
This paper is the documentation for a numerical code for quantum transport called KNIT. The KNIT library implements a generalization of the well known recursive Green function technique for a large class of multi-terminal mesoscopic systems with arbitrary geometries, topology and dimension. The systems are described by tight-biding Hamiltonians (with arbitrary internal degree of freedom like spin,…
▽ More
This paper is the documentation for a numerical code for quantum transport called KNIT. The KNIT library implements a generalization of the well known recursive Green function technique for a large class of multi-terminal mesoscopic systems with arbitrary geometries, topology and dimension. The systems are described by tight-biding Hamiltonians (with arbitrary internal degree of freedom like spin, electron/hole grading, orbitals...) and the calculation done within the non equilibrium Green function formalism (equivalent to the Landauer-Buttiker Scattering approach). KNIT can (and has) been applied to many different physical systems including quantum Hall effect, topological insulators, graphene ribbons, hybrid superconducting systems and ferromagnetic multilayers. KNIT main functionality is written in C++ and wrapped into Python, providing a simple and flexible interface for the user. Usual "input files" of numerical codes are replaced by small python scripts where the user builds the system and then computes whatever observable (conductance, noise, local density of states...) is needed. KNIT code can be obtained freely at the following url: http://inac.cea.fr/Pisp/xavier.waintal/KNIT.php
△ Less
Submitted 13 October, 2010;
originally announced October 2010.
