Perturbation theory, irrep truncations, and state preparation methods for quantum simulations of S U ( 3 ) SU(3) lattice gauge theory
Praveen Balaji, Cianán Conefrey-Shinozaki, Patrick Draper, Jason K. Elhaderi, Drishti Gupta, Luis Hidalgo, Andrew Lytle · 2025
Read Perturbation theory, irrep truncations, and state preparation methods for quantum simulations of S U ( 3 ) SU(3) lattice gauge theory by Praveen Balaji, Cianán Conefrey-Shinozaki, Patrick Draper, Jason K. Elhaderi, Drishti Gupta, Luis Hidalgo, Andrew Lytle (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-2509-25865.txt.
Introduction
The strong force's grid tale is a mark for new count gear, yet its ground state is hard to dress on such gear. Praveen Balaji and co-authors dress it in a 2025 study, in a charge first frame. Their cut of the frame's state space is graded by the drive in lone site knots, a fine grain past cuts.
Weak tie sums guide plain dress paths for the ground. Small grid tests, a flat four knot and a cube, weigh dress ways: slow drift, guess paths, and a wed of both. Faith and cost are set side by side. Two free tool packs ship with the work, for grid paths and for group weights.
Catalog record
| CREATOR | Praveen Balaji, Cianán Conefrey-Shinozaki, Patrick Draper, Jason K. Elhaderi, Drishti Gupta, Luis Hidalgo, Andrew Lytle |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 98 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-25865
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Perturbation theory, irrep truncations, and state preparation methods for quantum simulations of S U ( 3 ) SU(3) lattice gauge theory Praveen Balaji, Cianán Conefrey-Shinozaki, Patrick Draper, Jason K. Elhaderi, Drishti Gupta, Luis Hidalgo, Andrew Lytle Abstract We study methods for efficient preparation of approximate ground states of S U ( 3 ) SU(3) lattice gauge theory on quantum hardware. Working in a variant of the electric basis, we introduce a refinement of the irrep truncation based on the energy density of site singlets, which provides a finer gradation of simulation complexity. Using strong-coupling perturbation theory as a guide, we develop simple ansatz circuits for ground state preparation and test them via classical simulation on small lattices, including the 2 × 2 2\times 2 plaquette lattice in d = 2 d=2 and the cube in d = 3 d=3 . We contrast state fidelities and resource requirements of variational methods against adiabatic state preparation and introduce a method that hybridizes the two approaches. Finally, we report on the public release of ymcirc —a package of tools for building S U ( 3 ) SU(3) circuits and processing measurements—and pyclebsch , a package for efficiently computing S U ( N ) SU(N) Clebsch-Gordan coefficients. 1 Introduction A sufficiently advanced quantum computer could function as a unique and valuable tool for simulating quantum field theories (QFTs), providing access to nonperturbative phenomena in real time [ 1 , 2 , 3 , 4 ] . In particular, quantum simulations of lattice quantum chromodynamics (QCD) may be used to study parton distribution functions [ 5 , 6 ] , hadronization [ 7 , 8 ] , and classically inaccessible parts of the QCD phase diagram [ 9 , 10 ] , all of which are directly relevant for natural and laboratory phenomena. Schematically, these future quantum simulations will require encoding a truncation of the target Hilbert space onto the quantum computer, preparing an initial quantum state, time-evolving the quantum state according to an approximation of a target Hamiltonian, and measuring observables. Much effort has gone into developing this framework over the last decade, particularly focused on encoding and time evolution. Several approaches have been taken to encode and time-evolve lattice gauge theories (LGTs) on digital quantum hardware. The most common starting point is the Kogut-Susskind (KS) Hamiltonian [ 11 , 12 ] . Approaches differ, in part, in how bosonic gauge link degrees of freedom are truncated and encoded while addressing the Gauss law; a non-exhaustive list includes: working with eigenstates of the electric Hamiltonian (the electric basis), labeled by irreducible representations of a continuous gauge group [ 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 ] ; working with eigenstates of the magnetic Hamiltonian (the magnetic basis), labeled by group elements of a discrete gauge (sub)group [ 26 , 27 , 28 , 29 ] ; and working with a hybrid of both bases [ 30 , 31 , 32 , 33 ] . Although observables can be calculated after time evolution on any state, a large class of interesting objects – namely correlation functions in the ground state – require the addressing the state preparation problem. Thus, quantum algorithms that prepare ground states of LGTs are necessary. A full framework for simulating QFTs (and in particular calculating scattering amplitudes in \phi^{4} theory) was developed in [ 34 ] . There, state preparation was possible because the ground state of the free theory was known and directly preparable. Then, the ground state of the interacting theory was built with adiabatic state preparation (ASP), where the free-theory ground state is evolved with a “time-dependent” Hamiltonian whose couplings are slowly turned on over time. In the case of SU(N) LGT, a similar technique can be used; for example, the ground state at strong (weak) coupling is essentially the electric (magnetic) vacuum, and in principle the ground state at weaker (stronger) coupling can be prepared adiabatically from these starting points. However, accurate ASP may require quantum circuits that are too deep for noisy intermediate-scale quantum (NISQ) computers. An alternative approach, the variational quantum eigensolver (VQE) [ 35 , 36 ] , was developed as a hybrid classical-quantum algorithm that can iteratively construct a reasonable ground state with a relatively shallower circuit. The algorithm takes a reference state (which for SU(N) LGT can be, for example, the electric vacuum) that is evolved with a parameterized quantum circuit, called the ansatz. The energy of the output state is then measured, and a classical optimizer tunes the ansatz parameters; the process is iterated until convergence. VQE has been used to probe a diverse range of phenomena in LGTs [ 37 , 38 , 39 , 40 , 41 , 42 ] , including SU(3) LGT [ 43 ] . It has also seen an assortment of enhancements related to optimal ansatz choice [ 44 , 45 ] , scalability when increasing the lattice volume [ 46 , 47 , 48 , 49 ] , and adjustments to the ansatz based on the lattice volume [ 50 ] . With an eye towards the future goal of simulating of lattice QCD, in this work we explore methods for preparing ground states of SU(3) LGT in two and three spatial dimensions. We use both VQE and ASP, and we develop two hybrid state preparation methods: one where the VQE reference state at a given coupling is the output state from a VQE performed at a higher coupling, and another where ASP is initialized with VQE. We use strong-coupling perturbation theory (PT) to motivate our VQE ansätze. Although strong-coupling PT becomes unreliable at weak coupling, the target coupling for future simulations could be in a regime where PT is still a reasonable guide. We will explore this idea in some detail on small lattices and benchmark ground state energies and eigenstate fidelities against classical exact diagonalization (ED) of the KS Hamiltonian. Throughout this work we utilize the reduced electric basis of the KS Hamiltonian, described in [ 15 ] . This basis is an elaboration of the local multiplet basis of [ 51 , 52 , 14 ] , which leverages the gauge invariance of physical states to greatly reduce the original electric basis of [ 13 ] . Link degrees of freedom are described by SU(3) irreducible representations (irreps), and lattice sites carry degrees of freedom that describe the gauge-singlet fusion of the irreps meeting at the site. One issue we explore is how state preparation accuracy and cost are affected by a chosen irrep truncation. A simple truncation is to cut on the electric energy of links. Roughly this energy grows with irrep dimensionality, so a variant of this truncation is to cut on the number of tensor indices in the irrep. We label T_{r} the truncation that keeps up to all r -index tensor irreps. However, this truncation is extremely coarse, and small changes result in rapid variations in both the maximum accessible energy density and the Hilbert space dimensionality. To improve on it we introduce a bound B on the maximum allowed electric