pith. sign in

arxiv: 2606.27989 · v1 · pith:VZTQITSZnew · submitted 2026-06-26 · ✦ hep-lat

Diagonal Kenney-Laub Rational Approximation to the Overlap Operator using Wilson and Brillouin Kernel

Stephan Durr , Stylianos Gregoriou , Giannis Koutsou This is my paper

Pith reviewed 2026-06-29 02:04 UTC · model grok-4.3

classification ✦ hep-lat
keywords overlap operatorKenney-Laub iteratessign functionGinsparg-Wilson relationlattice QCDWilson fermionBrillouin fermion
0
0 comments X

The pith

Diagonal Kenney-Laub iterates give a more efficient approximation to the overlap Dirac operator sign function than Chebyshev polynomials.

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

The paper proposes approximating the sign function in the overlap Dirac operator using diagonal Kenney-Laub iterates rather than Chebyshev polynomials. These iterates need no initial spectral data on the kernel and lend themselves to partial fraction decomposition for fast evaluation. In numerical tests on quenched lattices, increasing the approximation order leads to smaller violations of the Ginsparg-Wilson relation and more stable critical quark masses, with the KL method showing gains in both symmetry preservation and speed over the polynomial baseline.

Core claim

Diagonal Kenney-Laub iterates can be used to approximate the matrix sign function in the overlap operator formulation, resulting in improved chiral symmetry and computational efficiency compared to the Chebyshev approach when implemented with Wilson or Brillouin kernels.

What carries the argument

The diagonal Kenney-Laub iterates applied to the sign function approximation, expressed through partial fraction decomposition.

If this is right

  • KL iterates reduce the violation of the Ginsparg-Wilson relation for given approximation order.
  • The critical bare quark mass shows improved behavior with rising order.
  • The method applies equally to Wilson and Brillouin kernel operators.
  • Partial fraction form enables practical and efficient implementation without spectral preprocessing.

Where Pith is reading between the lines

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

  • The approach may lower the overall cost of overlap fermion calculations in full QCD simulations.
  • Similar iterates could be tested in other contexts requiring accurate sign function approximations.
  • Scaling studies on larger volumes would clarify whether the efficiency gains persist.

Load-bearing premise

The advantages observed in a proof-of-concept on quenched lattices at one beta value hold more generally.

What would settle it

Demonstrating that Chebyshev polynomials achieve comparable or superior Ginsparg-Wilson preservation at equivalent computational cost on the same or finer lattices.

Figures

Figures reproduced from arXiv: 2606.27989 by Giannis Koutsou, Stephan Durr, Stylianos Gregoriou.

Figure 1
Figure 1. Figure 1: FIG. 1. Sign-squared violation [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Sign-squared violation [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Same data as in Fig. 2, plotted against the average computational cost (in core-hours per random vector). [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 2
Figure 2. Figure 2: These results demonstrate that (on the vast ma [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. PCAC mass [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5. Pion effective mass [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6. PCAC mass values, extracted from plateau fits to the [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: FIG. 7. PCAC mass (left panel) and pion mass squared (right panel) plotted against the overlap bare mass [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: FIG. 8. PCAC mass values plotted against the overlap bare mass, for the Wilson (left panel) and Brillouin (right panel) kernel, [PITH_FULL_IMAGE:figures/full_fig_p008_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: FIG. 9. Left: Negative critical bare masses ( [PITH_FULL_IMAGE:figures/full_fig_p009_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: FIG. 10. Same as Fig. 2, but for Chebyshev polynomial expansions of order [PITH_FULL_IMAGE:figures/full_fig_p009_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: FIG. 11. Same as Fig. 3, but augmented by the Chebyshev data. Results for two configurations (a well-conditioned one on the [PITH_FULL_IMAGE:figures/full_fig_p010_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: FIG. 12. PCAC mass values plotted against computational cost (in core-hours per spinor per configuration) at the overlap [PITH_FULL_IMAGE:figures/full_fig_p010_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: FIG. 13. Critical bare masses [PITH_FULL_IMAGE:figures/full_fig_p011_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: FIG. 14. Bar plot of the computational cost (in core-hours per [PITH_FULL_IMAGE:figures/full_fig_p011_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: FIG. 15. Stencil representations of the two ingredients of the Brillouin operator in four dimensions. Left: the x-direction [PITH_FULL_IMAGE:figures/full_fig_p014_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: FIG. 16. Exact scalar sign function (solid blue line) compared with approximations (dashed lines) using the Kenney-Laub [PITH_FULL_IMAGE:figures/full_fig_p015_16.png] view at source ↗
read the original abstract

We propose a formulation of the overlap Dirac operator in lattice QCD that employs diagonal Kenney-Laub (KL) iterates to approximate the matrix sign function. KL iterates require no prior spectral information about the kernel operator and, when expressed via their partial fraction decomposition, offer a practical and efficient approximation scheme. We evaluate this approach in a proof-of-concept implementation using quenched lattices at $\beta=6.2$ and two Dirac operator discretizations as a kernel, namely the Wilson and the Brillouin operators. By examining the approximate overlap operator's violation of the Ginsparg-Wilson relation and the critical bare quark mass for increasing approximation order, we find that KL iterates deliver enhanced chiral symmetry preservation and computational efficiency compared to the Chebyshev polynomial approach.

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 / 0 minor

Summary. The paper proposes a formulation of the overlap Dirac operator using diagonal Kenney-Laub iterates to approximate the matrix sign function, requiring no prior spectral information. It presents a proof-of-concept numerical evaluation on quenched SU(3) lattices at β=6.2 with Wilson and Brillouin kernels, comparing Ginsparg-Wilson relation violation and critical bare quark mass against Chebyshev polynomial approximations for increasing orders, and claims improved chiral symmetry preservation and computational efficiency.

Significance. If the reported advantages in approximation quality and efficiency are confirmed more broadly, the approach could provide a practical, parameter-free rational approximation scheme for overlap fermions that avoids spectral preconditioning steps common in other methods.

major comments (1)
  1. [Numerical results (as summarized in the abstract)] The central claim that KL iterates deliver enhanced chiral symmetry preservation and efficiency rests on comparisons performed exclusively on quenched ensembles at a single coupling (β=6.2). Spectral properties of the kernel operator can change with dynamical fermions or different β, so the observed gains may not persist; additional tests across a wider range of parameters are needed to support the generality of the conclusion.

Simulated Author's Rebuttal

1 responses · 1 unresolved

We thank the referee for their careful reading of the manuscript and for highlighting the scope of the numerical results. We address the major comment below.

read point-by-point responses

  1. Referee: The central claim that KL iterates deliver enhanced chiral symmetry preservation and efficiency rests on comparisons performed exclusively on quenched ensembles at a single coupling (β=6.2). Spectral properties of the kernel operator can change with dynamical fermions or different β, so the observed gains may not persist; additional tests across a wider range of parameters are needed to support the generality of the conclusion.

    Authors: We agree that the numerical evidence is restricted to quenched SU(3) ensembles at β=6.2, as stated in the abstract and Section 3 where the work is explicitly described as a proof-of-concept. All comparisons between KL iterates and Chebyshev polynomials are performed on identical configurations and kernels, so the reported improvements in Ginsparg-Wilson violation and critical mass are internally consistent within this setting. The manuscript does not claim universality beyond these ensembles. We will add a clarifying paragraph in the conclusions emphasizing the limited parameter range explored and the desirability of future tests with dynamical fermions and varied β. However, performing such additional simulations lies outside the present scope. revision: partial

standing simulated objections not resolved
  • Additional numerical tests on dynamical fermion ensembles and at different values of β to establish broader generality of the observed advantages.

Circularity Check

0 steps flagged

No circularity: claims rest on direct numerical measurements on external lattices

full rationale

The paper proposes diagonal Kenney-Laub iterates for sign-function approximation in the overlap operator and reports a proof-of-concept numerical study on quenched SU(3) lattices at β=6.2 using Wilson and Brillouin kernels. Comparisons of Ginsparg-Wilson violation and critical bare quark mass versus approximation order are obtained by explicit computation on those ensembles and contrasted with Chebyshev results; these are independent measurements, not quantities fitted to the target observables or defined in terms of the claimed improvements. No self-citations, uniqueness theorems, or ansatze are invoked as load-bearing steps in the abstract or described evaluation. The derivation chain consists of a standard rational approximation followed by lattice measurements whose outcomes are not forced by construction from the inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract provides no information to identify free parameters, axioms, or invented entities; all fields left empty.

pith-pipeline@v0.9.1-grok · 5660 in / 1085 out tokens · 52927 ms · 2026-06-29T02:04:30.214884+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

32 extracted references · 19 linked inside Pith

  1. [1]

    Neuberger, Exactly massless quarks on the lattice, Physics Letters B417, 141 (1998), hep-lat/9707022

    H. Neuberger, Exactly massless quarks on the lattice, Physics Letters B417, 141 (1998), hep-lat/9707022

    Pith/arXiv arXiv 1998

  2. [2]

    Neuberger, More about exactly massless quarks on the lattice, Physics Letters B427, 353 (1998), hep- lat/9801031

    H. Neuberger, More about exactly massless quarks on the lattice, Physics Letters B427, 353 (1998), hep- lat/9801031

    arXiv 1998

  3. [3]

    Neuberger, A Practical Implementation of the Overlap Dirac Operator, Physical Review Letters81, 4060 (1998), hep-lat/9806025

    H. Neuberger, A Practical Implementation of the Overlap Dirac Operator, Physical Review Letters81, 4060 (1998), hep-lat/9806025

    Pith/arXiv arXiv 1998

  4. [4]

    P. H. Ginsparg and K. G. Wilson, A remnant of chiral symmetry on the lattice, Physical Review D25, 2649 (1982)

    1982

  5. [5]

    L¨ uscher, Exact chiral symmetry on the lattice and the ginsparg-wilson relation, Physics Letters B428, 342 (1998), hep-lat/9802011

    M. L¨ uscher, Exact chiral symmetry on the lattice and the ginsparg-wilson relation, Physics Letters B428, 342 (1998), hep-lat/9802011

    Pith/arXiv arXiv 1998

  6. [6]

    L. D. Debbio and C. Pica, Topological susceptibility from the overlap, Journal of High Energy Physics (2003), hep- lat/0309145

    arXiv 2003

  7. [7]

    P. A. Boyle, L. Del Debbio, A. J¨ uttner, A. Khamseh, F. Sanfilippo, and J. T. Tsang, The decay constantsf D andf Ds in the continuum limit ofN f =2+1domain wall lattice QCD, JHEP12(12), 008, arXiv:1701.02644 [hep-lat]

    Pith/arXiv arXiv

  8. [8]

    Luscher, S

    M. Luscher, S. Sint, R. Sommer, and P. Weisz, Chiral symmetry and O(a) improvement in lattice QCD, Nucl. Phys. B478, 365 (1996), hep-lat/9605038

    Pith/arXiv arXiv 1996

  9. [9]

    R. G. Edwards, U. M. Heller, and R. Narayanan, A study of practical implementations of the overlap Dirac op- erator in four dimensions, Nuclear Physics B540, 457 (1999), hep-lat/9807017

    Pith/arXiv arXiv 1999

  10. [10]

    van den Eshof, A

    J. van den Eshof, A. Frommer, T. Lippert, K. Schilling and H. A. van der Vorst, Numerical methods for the QCD overlap operator. I. Sign-function and error bounds, Computer Physics Communications146, 203 (2002), hep-lat/0202025

    Pith/arXiv arXiv 2002

  11. [11]

    Kennedy, Approximation theory for matrices, Nuclear Physics B - Proceedings Supplements128, 107 (2004), hep-lat/0402037

    A. Kennedy, Approximation theory for matrices, Nuclear Physics B - Proceedings Supplements128, 107 (2004), hep-lat/0402037

    Pith/arXiv arXiv 2004

  12. [12]

    T. A. DeGrand (MILC), A Variant approach to the overlap action, Phys. Rev. D63, 034503 (2000), hep- lat/0007046

    arXiv 2000

  13. [13]

    Bietenholz, Convergence rate and locality of improved overlap fermions, Nucl

    W. Bietenholz, Convergence rate and locality of improved overlap fermions, Nucl. Phys. B644, 223 (2002), hep- lat/0204016

    arXiv 2002

  14. [14]

    Ikeda and S

    H. Ikeda and S. Hashimoto, O(a 2) improvement of the overlap-Dirac operator, PoSLA T2009, 082 (2009), arXiv:0912.4119 [hep-lat]

    Pith/arXiv arXiv 2009

  15. [15]

    Durr and G

    S. Durr and G. Koutsou, Brillouin improvement for Wilson fermions, Phys. Rev. D83, 114512 (2011), arXiv:1012.3615 [hep-lat]

    Pith/arXiv arXiv 2011

  16. [16]

    Y.-G. Cho, S. Hashimoto, A. J¨ uttner, T. Kaneko, M. Marinkovic, J.-I. Noaki, and J. T. Tsang, Improved lattice fermion action for heavy quarks, JHEP2015(5), 72, arXiv:1504.01630 [hep-lat]

    Pith/arXiv arXiv

  17. [17]

    Alexandru, M

    A. Alexandru, M. Lujan, C. Pelissier, B. Gamari, and F. X. Lee, Efficient implementation of the overlap op- erator on multi-GPUs, in2011 Symposium on Ap- plication Accelerators in High-Performance Computing (SAAHPC’11), IEEE Nucl.Sci.Symp.Conf.Rec. (2011) pp. 123–130, arXiv:1106.4964 [hep-lat]

    Pith/arXiv arXiv 2011

  18. [18]

    Brannick, A

    J. Brannick, A. Frommer, K. Kahl, B. Leder, M. Rottmann, and A. Strebel, Multigrid Precondition- ing for the Overlap Operator in Lattice QCD, Numer. Math.132, 463 (2016), arXiv:1410.7170 [hep-lat]

    Pith/arXiv arXiv 2016

  19. [19]

    Durr and G

    S. Durr and G. Koutsou, On the suitability of the Brillouin action as a kernel to the overlap procedure, arXiv:1701.00726 [hep-lat]

    Pith/arXiv arXiv

  20. [20]

    D¨ urr, S

    S. D¨ urr, S. Gregoriou, and G. Koutsou, Diagonal Kenney-Laub Rational Approximation to the Overlap Dirac Operator, in42th International Symposium on Lat- tice Field Theory(2025) arXiv:2512.20223 [hep-lat]

    arXiv 2025

  21. [21]

    Frommer, B

    A. Frommer, B. N¨ ockel, S. G¨ usken, T. Lippert, and K. Schilling, Many Masses on one Stroke: Economic Computation of Quark Propagators, International Jour- nal of Modern Physics C6, 627 (1995), hep-lat/9504020

    Pith/arXiv arXiv 1995

  22. [22]

    Jegerlehner, Krylov space solvers for shifted linear sys- tems, hep-lat/9612014

    B. Jegerlehner, Krylov space solvers for shifted linear sys- tems, hep-lat/9612014

    Pith/arXiv arXiv

  23. [23]

    C. S. Kenney and A. J. Laub, A hyperbolic tangent iden- tity and the geometry of pad´ e sign function iterations, Numerical Algorithms7, 111 (1994)

    1994

  24. [24]

    Albaneseet al.(APE), Glueball Masses and String Tension in Lattice QCD, Phys

    M. Albaneseet al.(APE), Glueball Masses and String Tension in Lattice QCD, Phys. Lett. B192, 163 (1987)

    1987

  25. [25]

    Cundy, A

    N. Cundy, A. D. Kennedy, and A. Schafer, A lattice Dirac operator for QCD with light dynamical quarks, Nucl. Phys. B845, 30 (2011), arXiv:1010.5629 [hep-lat]

    Pith/arXiv arXiv 2011

  26. [26]

    J¨ ulich Supercomputing Centre, JURECA: Data Centric and Booster Modules implementing the Modular Super- computing Architecture at J¨ ulich Supercomputing Cen- tre, Journal of large-scale research facilities7(2021)

    2021

  27. [27]

    Neuberger, Bounds on the wilson dirac operator, Physical Review D61, 085015 (1999), hep-lat/9911004

    H. Neuberger, Bounds on the wilson dirac operator, Physical Review D61, 085015 (1999), hep-lat/9911004

    Pith/arXiv arXiv 1999

  28. [28]

    S. Durr, G. Koutsou, and T. Lippert, Meson and Baryon dispersion relations with Brillouin fermions, Phys. Rev. D86, 114514 (2012), arXiv:1208.6270 [hep-lat]

    Pith/arXiv arXiv 2012

  29. [29]

    Durr, Portable CPU implementation of Wilson, Bril- louin and Susskind fermions in lattice QCD, Comput

    S. Durr, Portable CPU implementation of Wilson, Bril- louin and Susskind fermions in lattice QCD, Comput. Phys. Commun.282, 108555 (2023), arXiv:2112.14640 [hep-lat]

    arXiv 2023

  30. [30]

    Kenney and A

    C. Kenney and A. J. Laub, Rational Iterative Methods for the Matrix Sign Function, SIAM Journal on Matrix Analysis and Applications12, 273 (1991)

    1991

  31. [31]

    Higham,Functions of Matrices: Theory and Compu- tation(Society for Industrial and Applied Mathematics, 2008)

    N. Higham,Functions of Matrices: Theory and Compu- tation(Society for Industrial and Applied Mathematics, 2008)

    2008

  32. [32]

    Press,Numerical Recipes 3rd Edition: The Art of Sci- entific Computing(Cambridge University Press, 2007)

    W. Press,Numerical Recipes 3rd Edition: The Art of Sci- entific Computing(Cambridge University Press, 2007)

    2007