pith. sign in

arxiv: 2511.14914 · v3 · submitted 2025-11-18 · 🪐 quant-ph

Exact Factorization of Unitary Transformations with Spin-Adapted Generators

Paarth Jain , Artur F. Izmaylov , Erik R. Kjellgren This is my paper

Pith reviewed 2026-05-17 20:08 UTC · model grok-4.3

classification 🪐 quant-ph
keywords spin symmetryvariational quantum algorithmsPauli operatorsLie algebrasunitary factorizationquantum simulationmolecular electronic structure
0
0 comments X

The pith

Spin-adapted unitaries factor exactly into ordered products of Pauli exponentials.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper shows how to rewrite unitaries that preserve spin symmetry, coming from fermionic double excitation rotations, as sequences of Pauli operator exponentials. This matters for variational quantum algorithms because it lets hardware implementations keep physical spin symmetry by construction while simulating molecular electronic states. The method works by noting that the basic operators inside these generators close into small Lie algebras, then recasts the entire factorization task as a low-dimensional numerical optimization problem solved in the adjoint representation. Solving the optimization yields an exact reparametrization that matches the original unitary without any symbolic manipulation. The resulting circuits therefore cost fewer gates and automatically respect spin when used inside variational loops for chemistry.

Core claim

We introduce an exact and computationally efficient factorization of spin-adapted unitaries derived from fermionic double excitation and deexcitation rotations. These unitaries are expressed as ordered products of exponentials of Pauli operators. Our method exploits the fact that the elementary operators in these generators form small Lie algebras. By working in the adjoint representation of these algebras, we reformulate the factorization problem as a low-dimensional nonlinear optimization over matrix exponentials. This approach enables precise numerical reparametrization of the unitaries without relying on symbolic manipulations.

What carries the argument

Reformulation of the factorization as low-dimensional nonlinear optimization over matrix exponentials in the adjoint representation of the small Lie algebras formed by the elementary operators.

If this is right

  • The factorization supplies a practical route to symmetry-conserving quantum circuits inside variational algorithms.
  • Spin symmetry is preserved by design in every implemented transformation.
  • Gate count and implementation cost on quantum hardware are reduced compared with generic decompositions.
  • Electronic states in molecular simulations remain accurately represented without symmetry-breaking errors.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same Lie-algebra reduction might apply directly to other conserved quantities such as particle number or point-group symmetry.
  • Embedding the numerical optimizer inside existing variational quantum eigensolver loops could shrink the total parameter count while keeping symmetry exact.
  • Because the optimization dimension stays small, the technique may remain tractable even when the underlying molecule grows beyond current exact-diagonalization limits.

Load-bearing premise

The elementary operators inside the spin-adapted generators close into small Lie algebras that allow the factorization to be recast as a low-dimensional nonlinear optimization in the adjoint representation.

What would settle it

Apply the optimized product of Pauli exponentials to a simple reference state for a concrete double-excitation generator and verify that the result agrees with direct application of the original fermionic unitary to machine precision.

Figures

Figures reproduced from arXiv: 2511.14914 by Artur F. Izmaylov, Erik R. Kjellgren, Paarth Jain.

Figure 1
Figure 1. Figure 1: shows the convergence of adaptive VQE calculations for H2O relative to the full configuration interaction (FCI) energy. Using the fermionic operator pool (blue dashed line) required 104 variational parameters to reach the gradient threshold of 10−5 a.u. In contrast, the spin-adapted fermionic operator pool (red dashed line) achieved convergence with only 54 parameters, demonstrating that enforcing total sp… view at source ↗
Figure 2
Figure 2. Figure 2: presents analogous results for BeH2. The same qualitative trend is observed: the fermionic operator pool required 95 parameters to converge, whereas the spin-adapted fermionic pool reached convergence with only 55 parameters. Again, the SSADpair operator pool became trapped in a local minimum and failed to reach the FCI energy. These results confirm that incorporating total spin symmetry systematically red… view at source ↗
Figure 3
Figure 3. Figure 3: displays the potential energy curve of O2 for bond lengths between 1.8 and 2.4 Å. When using the spin-adapted fermionic operator pool (SSADSA, red circles), the adaptive VQE energies follow the FCI triplet curve (black line) across all bond lengths. In contrast, using the conventional fermionic operator pool (SD, blue diamonds) leads to a collapse to the lower-energy singlet solution (green curve), even wh… view at source ↗
read the original abstract

Preserving spin symmetry in variational quantum algorithms is essential for producing physically meaningful electronic wavefunctions. Implementing spin-adapted transformations on quantum hardware, however, is challenging because the corresponding fermionic generators translate into noncommuting Pauli operators. In this work, we introduce an exact and computationally efficient factorization of spin-adapted unitaries derived from fermionic double excitation and deexcitation rotations. These unitaries are expressed as ordered products of exponentials of Pauli operators. Our method exploits the fact that the elementary operators in these generators form small Lie algebras. By working in the adjoint representation of these algebras, we reformulate the factorization problem as a low-dimensional nonlinear optimization over matrix exponentials. This approach enables precise numerical reparametrization of the unitaries without relying on symbolic manipulations. The proposed factorization provides a practical strategy for constructing symmetry-conserving quantum circuits within variational algorithms. It preserves spin symmetry by design, reduces implementation cost, and ensures the accurate representation of electronic states in quantum simulations of molecular systems.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The manuscript claims to introduce an exact factorization of spin-adapted unitaries arising from fermionic double-excitation and de-excitation rotations. These unitaries are expressed as ordered products of Pauli exponentials by exploiting the closure of the elementary operators into small Lie algebras, recasting the factorization as a low-dimensional nonlinear optimization problem over matrix exponentials in the adjoint representation. The approach is positioned as a practical tool for constructing symmetry-preserving circuits in variational quantum algorithms for molecular simulations.

Significance. If the central claim holds, the work supplies a concrete route to implement spin-adapted transformations on quantum hardware without Trotterization or approximation, directly supporting symmetry-conserving variational quantum eigensolvers. The emphasis on exactness via algebraic closure and numerical reparametrization in a low-dimensional adjoint space is a potentially useful addition to the toolkit for quantum chemistry simulations.

major comments (1)
  1. [Abstract and §2] The abstract asserts that the factorization is exact because the generators close into small Lie algebras, yet the manuscript provides neither an explicit derivation of the adjoint equations nor a worked example showing that the nonlinear optimization recovers the original unitary to machine precision. This verification step is load-bearing for the central claim of exactness.
minor comments (2)
  1. [§3] Notation for the adjoint representation and the objective function of the optimization should be introduced with a small concrete example (e.g., a two-orbital double excitation) before the general case.
  2. [§4] The manuscript would benefit from a brief comparison, even schematic, of gate counts or circuit depth against standard Trotterized or qubit-excitation implementations.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their positive assessment of the work's significance and for the constructive recommendation. We address the single major comment below and have revised the manuscript to incorporate the requested verification details.

read point-by-point responses

  1. Referee: [Abstract and §2] The abstract asserts that the factorization is exact because the generators close into small Lie algebras, yet the manuscript provides neither an explicit derivation of the adjoint equations nor a worked example showing that the nonlinear optimization recovers the original unitary to machine precision. This verification step is load-bearing for the central claim of exactness.

    Authors: We agree that an explicit derivation of the adjoint equations and a concrete numerical example are necessary to fully substantiate the exactness claim. The original manuscript describes the reformulation in the adjoint representation but does not present the equations in full detail or include a verification example. In the revised manuscript we have expanded Section 2 with a complete step-by-step derivation of the adjoint-action equations for the relevant small Lie algebras (so(3) and su(2) subalgebras arising from spin-adapted double excitations). We have also added a new subsection containing a worked numerical example for a representative double-excitation generator, in which the low-dimensional nonlinear optimization recovers the original unitary to machine precision (Frobenius-norm residual < 10^{-14}). These additions directly address the verification step identified by the referee. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation is self-contained

full rationale

The paper's factorization is obtained by noting that the relevant fermionic generators close into small Lie algebras (a standard, externally verifiable algebraic fact) and then recasting the unitary factorization as an exact low-dimensional nonlinear optimization over matrix exponentials in the adjoint representation. This step follows directly from the definition of the adjoint representation and the Baker-Campbell-Hausdorff relations within a finite-dimensional algebra; it does not redefine the target unitary in terms of its own outputs, fit parameters to the result being derived, or rely on load-bearing self-citations. The construction therefore remains independent of the specific numerical values or final circuit form it produces.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the domain assumption that the relevant fermionic generators close into small Lie algebras whose adjoint representation allows numerical reparametrization.

axioms (1)
  • domain assumption Elementary operators in the spin-adapted generators form small Lie algebras
    Invoked to justify reformulating the factorization as low-dimensional nonlinear optimization over matrix exponentials.

pith-pipeline@v0.9.0 · 5474 in / 1144 out tokens · 51489 ms · 2026-05-17T20:08:40.575215+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

16 extracted references · 16 canonical work pages

  1. [1]

    θ KX i=1 ciAi # = exp

    Because both factors have spectra{±1, 0}, the total spectrum of each Lie algebra generator is also{±1,0}. For example, one basis of the seniority-2 Lie algebra is A1 = ˆG aαbβ iαiβ ,(7) A2 = ˆG aβ bα iαiβ ,(8) A3 = (1−ˆn(iα)−ˆn(iβ)) ˆG aβ bα aαbβ ,(9) A4 =− 1−ˆn(aα)−ˆn(bβ) + 2ˆn(aα)ˆn(bβ) ˆG aβ bα iαiβ ,(10) A5 = 1−ˆn(aβ)−ˆn(bα) + 2ˆn(aβ)ˆn(bα) ˆG aαbβ iα...

    work page

  2. [2]

    A quantum computing view on unitary coupled cluster theory.Chem

    A Anand, P Schleich, S Alperin-Lea, PWK Jensen, S Sim, M Díaz-Tinoco, JS Kottmann, M Degroote, AF Izmaylov, and A Aspuru-Guzik. A quantum computing view on unitary coupled cluster theory.Chem. Soc. Rev., 51:1659–1684, 2022

    work page 2022

  3. [3]

    Ryabinkin, Tzu-Ching Yen, Scott N

    Ilya G. Ryabinkin, Tzu-Ching Yen, Scott N. Genin, and Artur F. Izmaylov. Qubit coupled- cluster method: A systematic approach to quantum chemistry on a quantum computer.J. Chem Theory Comput., 14:6317–6326, 2018

    work page 2018

  4. [4]

    Iterative qubit coupled cluster approach with efficient screening of generators.J

    IG Ryabinkin, RA Lang, SN Genin, and AF Izmaylov. Iterative qubit coupled cluster approach with efficient screening of generators.J. Chem. Theory Comput., 16(2):1055–1063, 2020

    work page 2020

  5. [5]

    An adaptive variational algorithm for exact molecular simulations on a quantum computer.Nat

    Harper R Grimsley, Sophia E Economou, Edwin Barnes, and Nicholas J Mayhall. An adaptive variational algorithm for exact molecular simulations on a quantum computer.Nat. Commun., 10(1):3007, July 2019

    work page 2019

  6. [6]

    Shkolnikov, George S

    Ho Lun Tang, V.O. Shkolnikov, George S. Barron, Harper R. Grimsley, Nicholas J. Mayhall, Ed- win Barnes, and Sophia E. Economou. Qubit-adapt-vqe: An adaptive algorithm for constructing hardware-efficient ansätze on a quantum processor.PRX Quantum, 2(2), April 2021

    work page 2021

  7. [7]

    Hugh G. A. Burton, Daniel Marti-Dafcik, David P. Tew, and David J. Wales. Exact electronic states with shallow quantum circuits from global optimisation.npj Quantum Information, 9(1), July 2023

    work page 2023

  8. [8]

    Constrained variational quantum eigensolver: Quantum computer search engine in the fock space.J

    Ilya G Ryabinkin, Scott N Genin, and Artur F Izmaylov. Constrained variational quantum eigensolver: Quantum computer search engine in the fock space.J. Chem. Theory Comp., 15:249–255, 2019

    work page 2019

  9. [9]

    Lang, and Artur F

    Tzu-Ching Yen, Robert A. Lang, and Artur F. Izmaylov. Exact and approximate symmetry projectors for the electronic structure problem on a quantum computer. J. Chem. Phys., 151:164111, 2019

    work page 2019

  10. [10]

    Local, expressive, quantum-number-preserving vqe ansätze for fermionic systems.New Journal of Physics, 23(11):113010, nov 2021

    Gian-Luca R Anselmetti, David Wierichs, Christian Gogolin, and Robert M Parrish. Local, expressive, quantum-number-preserving vqe ansätze for fermionic systems.New Journal of Physics, 23(11):113010, nov 2021. 14

    work page 2021

  11. [11]

    Ðoković Dragomir Ž and Karl H. Hofmann. The surjectivity question for the exponential function of real lie groups: A status report.Journal of Lie Theory, 7:171–199, 1997

    work page 1997

  12. [12]

    SpinAdapted.https://github.com/paarth7777/SpinAdapted, 2025

    Paarth Jain. SpinAdapted.https://github.com/paarth7777/SpinAdapted, 2025

    work page 2025

  13. [13]

    Kjellgren and K

    E. Kjellgren and K. M. Ziems. SlowQuant.https://github.com/erikkjellgren/SlowQuant, 2025

    work page 2025

  14. [14]

    Oliphant, Matt Haberland, Tyler Reddy, David Cournapeau, Evgeni Burovski, Pearu Peterson, Warren Weckesser, Jonathan Bright, Stéfan J

    Pauli Virtanen, Ralf Gommers, Travis E. Oliphant, Matt Haberland, Tyler Reddy, David Cournapeau, Evgeni Burovski, Pearu Peterson, Warren Weckesser, Jonathan Bright, Stéfan J. van der Walt, Matthew Brett, Joshua Wilson, K. Jarrod Millman, Nikolay Mayorov, Andrew R. J. Nelson, Eric Jones, Robert Kern, Eric Larson, C J Carey, İlhan Polat, Yu Feng, Eric W. Mo...

    work page 2020

  15. [15]

    Nakagawa

    Kohdai Kuroiwa and Yuya O. Nakagawa. Penalty methods for a variational quantum eigensolver. Physical Review Research, 3(1), February 2021

    work page 2021

  16. [16]

    partners

    Takashi Tsuchimochi, Yuto Mori, and Seiichiro L. Ten-no. Spin-projection for quantum computation: A low-depth approach to strong correlation.Physical Review Research, 2(4), October 2020. 15 Appendix A: Seniority 4, intermediate singlet This algebra is generated byA1 = ˆG aαbβ iαjβ , A2 = ˆG aβ bα iβ jα, A3 = ˆG aαbβ iβ jα, A4 = ˆG aβ bα iαjβ from the symm...

    work page 2020