Tensor-network toolbox for probing dynamics of non-Abelian gauge theories

Indrakshi Raychowdhury · 2025

Read Tensor-network toolbox for probing dynamics of non-Abelian gauge theories by Indrakshi Raychowdhury (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-18301.txt.

Introduction

A team including Indrakshi Raychowdhury wrote a physics paper in two thousand twenty-five on simulating forces inside matter with tensor networks. The forces are gauge theories, the framework behind the strong nuclear force. Such systems are hard to simulate because standard methods hit a wall known as the sign problem.

Tensor networks sidestep that wall by tracking the system through a chain of linked tensors. The team built and tested a version for a theory with the SU(2) symmetry group, using a formulation based on loops, strings, and hadrons. They report progress on both still and moving measures of the theory in one space dimension.

Catalog record

CREATORIndrakshi Raychowdhury
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE21 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-18301

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Tensor-network toolbox for probing dynamics of non-Abelian gauge theories Indrakshi Raychowdhury Abstract Tensor-network methods enable probing dynamics of strongly interacting quantum many-body systems, including gauge theories, via Hamiltonian simulation, hence bypassing sign problems. They also have the potential to inform efficient quantum-simulation algorithms of the same theories. We develop and benchmark a matrix-product-state ansatz for the SU(2) lattice gauge theory using the loop-string-hadron formulation. This formulation has been demonstrated to be advantageous in Hamiltonian simulation of non-Abelian gauge theories. It is applicable to both SU(2) and SU(3) gauge groups, to periodic and open boundary conditions, and to 1+1 and higher dimensions. In this work, we report on progress in computing static and dynamical observables in a SU(2) gauge theory in (1+1)D, pushing the boundary of existing studies. [a,b]Emil Mathew 1 Introduction Tensor networks (TNs) are promising sign-problem-free tools that enable studies of static and dynamic properties of quantum many-body systems [ 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 ] . In the context of high-energy and nuclear physics, much effort has been directed at lattice gauge theories (LGTs), both Abelian [ 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 ] and non-Abelian gauge theories [ 24 , 25 , 26 , 27 , 28 , 29 ] , in 1+1 and higher dimensions [ 30 , 31 , 32 , 33 , 34 , 35 , 36 ] , see also Refs. [ 37 , 8 , 38 , 39 ] for reviews. This work reports a tensor-network study of the SU(2) LGT in (1+1)D, employing a recently-developed theoretical formulation [ 40 ] . This LGT has been extensively studied in Refs. [ 24 , 25 , 26 ] , where different formulations of the SU(2) gauge theory [ 41 ] , namely truncated angular-momentum basis and purely fermionic basis, were used to compute static and dynamic quantities. The static quantities of interest comprise ground-state energy estimation via the density-matrix renormalization group (DMRG) and its continuum-limit extrapolation for various model parameters. In contrast, the dynamical computation explores string-breaking dynamics in the presence of static and dynamical charges on the lattice. (See Refs. [ 42 , 43 , 44 , 45 , 46 , 47 ] for recent quantum-simulation experiments of string breaking in Abelian models). While Refs. [ 24 , 26 ] observe string-breaking dynamics in the SU(2) LGT, they are either constrained to small Hilbert-space truncation cutoffs and lattice sizes (Ref. [ 24 ] ) or to the evolution of only static charges (Ref. [ 26 ] ). This work is aimed at extending existing results, setting the stage for studies of other non-equilibrium processes, such as post-collision phenomena, in the SU(2) gauge theory. Specifically, we simulate the dynamics of string breaking by exploring sufficiently long strings composed of dynamical fermions embedded in sufficiently large lattice volumes to enable approaching the continuum limit. We also control the gauge-boson-truncation effects, reaching larger cutoffs than previously accessible. As a result, we observe richer phenomenology compared to previous studies. Our work is enabled by a gauge-invariant reformulation of the Kogut-Susskind LGT [ 48 ] based on loop, string, and hadron degrees of freedom [ 40 ] , resulting in a simpler Abelianized theory. This proceedings is organized as follows: Section 2 briefly reviews the loop-string-hadron formalism, followed by Sec. 3 , where the tensor-network framework for the LSH formulation is introduced. In Sec. 4 , we compute the ground-state energy along with the effect of external static charges, i.e., the static potential. Section 5 contains a study of the quenched dynamics of dynamical strings and the string-breaking phenomenon. We conclude in Sec. 6 . 2 Loop-string-hadron formulation The loop-string-hadron (LSH) formulation replaces the gauge-invariant degrees of freedom, such as Wilson loops/lines, by local snapshots of the same. These snapshots are defined using a gauge-invariant and orthonormal LSH basis, which in (1+1)D is characterized by a set of three integers \{n_{l},n_{i},n_{o}\} . n_{l}\in\mathbb{Z}^{+}\cup\{0\} is a bosonic quantum number while n_{i},n_{o}\in\{0,1\} are fermionic quantum numbers. The global state is a tensor-product state associated with these quantum numbers at each site r of the spatial lattice, i.e., \bigotimes_{r=1}^{N}|n_{l},n_{i},n_{o}\rangle_{r} , glued together via an Abelian Gauss law (AGL), N_{L}(r)=N_{R}(r+1) , where N_{L}(r)\coloneq\big[n_{l}+n_{o}(1-n_{i})\big]\Big|_{r} and N_{R}(r)\coloneq\big[n_{l}+n_{i}(1-n_{o})\big]\Big|_{r} , \forall r . For a more detailed exposition of the LSH formalism, see Ref. [ 40 ] . The Hamiltonian in the LSH formulation can be written as: \displaystyle\hat{H}^{(\rm LSH)} \displaystyle=\frac{g^{2}a}{4}\sum_{r=1}^{N-1}\Bigg[\frac{\hat{N}_{L}(r)}{2}\left(\frac{\hat{N}_{L}(r)}{2}+1\right)+\Big(\hat{N}_{L}(r)\leftrightarrow\hat{N}_{R}(r+1)\Big)\Bigg]+m\sum_{r=1}^{N}(-1)^{r}(\hat{n}_{i}(r)+\hat{n}_{o}(r)) \displaystyle+\frac{1}{2a}\sum_{r=1}^{N-1}\Bigg\{\frac{1}{\sqrt{\hat{N}_{L}(r)+1}}\Big[\hat{\mathcal{S}}_{\text{out}}^{++}(r)\hat{\mathcal{S}}_{\text{in}}^{+-}(r+1)+\hat{\mathcal{S}}_{\text{out}}^{+-}(r)\hat{\mathcal{S}}_{\text{in}}^{--}(r+1)\Big]\frac{1}{\sqrt{\hat{N}_{R}(r+1)+1}}+\text{H.c}.\Bigg\}. (1) Here, g is the coupling strength, m is the mass of the staggered fermions, and a is the lattice spacing. The operators used in the Hamiltonian are defined in Table 1 . In this work, we apply open boundary conditions (OBCs) but our formalism is also applicable to periodic boundary conditions (PBCs). Table 1: The list of relevant operators for defining the LSH Hamiltonian and dynamical strings in (1+1)D. Note that \hat{\lambda}^{-}|0,n_{i},n_{o}\rangle=0 and the symbol \oplus is used to denote addition modulo 2. These definitions are valid for a single site; when embedded in a lattice, the operators carry a position label r and the \chi(r) definitions must be endowed with additional structure to ensure proper anticommutation relations [ 40 ] . 3 Tensor-network framework for the LSH formulation A matrix-product-state (MPS) ansatz for a (1+1)D system with N spatial sites is given by |{\Psi[A]}\rangle=\sum_{p_{1},\ldots,p_{N}}\sum_{a_{1},\ldots,a_{N-1}}A^{a_{1}}_{p_{1}}A^{a_{1},a_{2}}_{p_{2}}\ldots A^{a_{N-1}}_{p_{N}}|p_{1},p_{2},\ldots,p_{N}\rangle. Here, p_{r} and a_{r} label the physical and virtual degrees of freedom at site r , respectively. The maximum number of values of a_{r} is referred to as the bond dimension. In order to represent LSH states using an MPS, the infinite-dimensional Hilbert space associated with the gauge bosons must be truncated. One could either restrict the loop quantum number to n_{l}(r)\leq n_{l,\mathrm{max}} , \forall r , leading to a local physical Hilbert-space dimension of 4(n_{l,\mathrm{max}}+1) . The other choice is to restrict the value of the N_{L/R} quantum number to N_{L/R}\leq 2J_{\rm max} , where J_{\text{max}} is the cutoff on the irreducible representation of the gauge group in the Kogut-Susskind formulation. The truncated LSH state in either case can be represented with an MPS, with |p_{r},\ldots,p_{N}\rangle\equiv\bigotimes_{r=1}^{N}|p_{r}\rangle and |p_{r}\rangle\equiv|n_{l}(r),n_{i}(r),n_{o}(r)\rangle . The LSH Hamiltonian admits two U(1) global symmetries resulting in the conservation of charges Q=\sum_{r=1}^{N}\big[n_{i}(r)+n_{o}(r)\big] and q=\sum_{r=1}^{N}\big[n_{o}(r)-n_{i}(r)\big] . For OBCs, Q\in[0,2N] and q\in[-N,N] , and only certain combinations of Q and q are consistent with the AGL. We restrict ourselves to the q=0 and Q=N sector. These global symmetries endow the local tensors with a block-diagonal structure, simplifying algorithmic complexity. Additionally, we impose a penalty term in the Hamiltonian to ensure the satisfaction of the AGL. The LSH Hamiltonian comprises at most nearest-neighbor interactions, allowing a compact matrix-product-operator (MPO) representation. Finally, tensor-network techniques rely upon truncating the bond dimension at some maximum value, denoted as D_{\rm max} , to enable efficient computations. The tensors A constitute the variational degrees of freedom which are optimized to approximate ground states and perform time evolution. We rely upon existing methods defined in the ITensors.jl package [ 49 ] to construct and optimize the MPS and MPOs needed for this study. 4 Probing static properties Using standard

Read the complete text