Forward-mode automatic differentiation for the tensor renormalization group and its relation to the impurity method

Abstract

We propose a forward-mode automatic differentiation (AD) framework for tensor renormalization group methods. In this approach, evaluating the derivatives of the partition function up to the order of $k$ increases the matrix-multiplication cost by a factor of $(k+1)(k+2)/2$ compared to computing the free energy alone, and the memory footprint is only $k+1$ times that of the original calculation. In the limit where the derivatives of the singular value decomposition are neglected, we establish a theoretical correspondence between our forward-mode AD and conventional impurity methods. Numerically, we find that the proposed AD algorithm can calculate internal energy and specific heat significantly higher accuracy than the impurity method at comparable computational cost. We also provide a practical procedure to extract critical exponents from derivatives of the renormalized tensor in tensor renormalization group calculations in both two and three dimensions. In addition, we discuss how to efficiently differentiate an arbitrary tensor network.

24 pages, 8 figures

References in corpus (60)

• Tensor-Entanglement-Filtering Renormalization Approach and Symmetry Protected Topological Order
• Tensor renormalization group approach to 2D classical lattice models
• Precision Islands in the Ising and $O(N)$ Models
• Coarse-graining renormalization by higher-order singular value decomposition
• Tensor Network Renormalization
• Differentiable Programming Tensor Networks
• Loop optimization for tensor network renormalization
• Grassmann Tensor Renormalization Group Approach to One-Flavor Lattice Schwinger Model
• Renormalization of tensor networks using graph independent local truncations
• Algorithms for tensor network renormalization
• Tensor renormalization group study of classical XY model on the square lattice
• Critical behavior of lattice Schwinger model with topological term at $\theta =\pi$ using Grassmann tensor renormalization group
• Anisotropic Tensor Renormalization Group
• Grassmann tensor renormalization group for one-flavor lattice Gross-Neveu model with finite chemical potential
• Tensor renormalization group study of the non-Abelian Higgs model in two dimensions
• Phase transition of four-dimensional Ising model with higher-order tensor renormalization group
• Higher order tensor renormalization group for relativistic fermion systems
• Tensor renormalization group analysis of CP(N-1) model
• Calculation of fermionic Green functions with Grassmann higher-order tensor renormalization group
• Phase Transitions of Ferromagnetic Potts Models on the Simple Cubic Lattice
• Tensor renormalization group approach to four-dimensional complex ${\varphi }^4$ theory at finite density
• Automatic Differentiation for Second Renormalization of Tensor Networks
• Calculation of higher-order moments by higher-order tensor renormalization group
• Tensor Renormalization Group with Randomized Singular Value Decomposition
• Boundary Tensor Renormalization Group
• Finite-size and finite bond dimension effects of tensor network renormalization
• Improving the efficiency of variational tensor network algorithms
• Computing the renormalization group flow of two-dimensional ${\varphi }^4$ theory with tensor networks
• Restoration of chiral symmetry in cold and dense Nambu--Jona-Lasinio model with tensor renormalization group
• Tensor renormalization group and the volume independence in 2D U($N$) and SU($N$) gauge theories
• Tensor renormalization group study of the 3d $O(2)$ model
• Accurate simulation of q-state clock model
• Logarithmic finite-size scaling correction to the leading Fisher zeros in the p-state clock model: A higher-order tensor renormalization group study
• Tensor renormalization group study of (3+1)-dimensional $Z_2$ gauge-Higgs model at finite density
• Grassmann higher-order tensor renormalization group approach for two-dimensional strong-coupling QCD
• Toward tensor renormalization group study of three-dimensional non-Abelian gauge theory
• Critical endpoint of (3+1)-dimensional finite density $Z_3$ gauge-Higgs model with tensor renormalization group
• Phase transition of four-dimensional lattice ${\varphi }^4$ theory with tensor renormalization group
• Nuclear norm regularized loop optimization for tensor network
• Tensor Renormalization Group Algorithms with a Projective Truncation Method
• Phase structure of the CP(1) model in the presence of a topological $\theta$-term
• Tensor renormalization group approach to (1+1)-dimensional Hubbard model
• Boundary conformal spectrum and surface critical behaviors of the classical spin systems: a tensor network renormalization study
• Global optimization of tensor renormalization group using the corner transfer matrix
• Phase diagram of generalized XY model using tensor renormalization group
• Stable and efficient differentiation of tensor network algorithms
• Tensor renormalization group approach to (1+1)-dimensional SU(2) principal chiral model at finite density
• Multi-impurity method for the bond-weighted tensor renormalization group
• Improved coarse-graining methods on two dimensional tensor networks including fermions
• Tensor renormalization group study of (1+1)-dimensional U(1) gauge-Higgs model at $\theta =\pi$ with Lüscher's admissibility condition
• Calculation of critical exponents on fractal lattice Ising model by higher-order tensor renormalization group method
• Phase structure of (3+1)-dimensional dense two-color QCD at $T=0$ in the strong coupling limit with the tensor renormalization group
• Initial tensor construction and dependence of the tensor renormalization group on initial tensors
• Spectroscopy with the tensor renormalization group method
• Tensor Renormalization Group Calculations of Partition-Function Ratios
• Phase structure analysis of 2d lattice CP(1) model with $\theta$ term using tensor renormalization group method
• Tensor renormalization group approach to critical phenomena via symmetry-twisted partition functions
• Tensor network renormalization approach to antiferromagnetic 6-state clock model on the Union Jack lattice
• Application of tensor network method to two dimensional lattice $\mathcal{N}=1$ Wess-Zumino model
• Variational boundary based tensor network renormalization group

Cited by in corpus (1)

• A Practical Introduction to Tensor Network Renormalization with TNRKit.jl