REVIEW
Sign of Wilson-Fermion Determinant using Contour-Integral Spectral Projection
High Energy Physics - Lattice · Submitted 23 Sep 2026 · 32 references
Pith's verdict: Genuinely useful contour projection for the Wilson determinant sign, verified on the tested CLS ensembles — but the completeness claim rides on a certified left boundary that is thinner than the prose suggests.
T0 · AI review
Why Pith says this
Highlight any text to comment on it.
Referee report
The paper
The paper claims
The sign of a Wilson fermion determinant can be read from a contour integral that counts its negative-real modes, reproducing all of them to machine precision on CLS ensembles.
The reading
What carries the argument
The load-bearing object is the Riesz spectral projector $P_\Gamma = \frac{1}{2\pi i}\oint_\Gamma (zI-\widehat{D})^{-1}\,dz$, discretized with $N_q$ quadrature nodes into $\widehat{P}_\Gamma \simeq \sum_j \omega_j (z_j I - \widehat{D})^{-1}$, which projects onto the subspace spanned by eigenvalues inside an ellipse $\Gamma$ spanning the real interval $[X_L, 10^{-4}]$. Three pieces carry the argument: a shifted-DFL-SAP-GCR solver, adapted from openQCD to act on the shifted even Schur complement $\widehat{D}-zI$ rather than the full $D$, solves each shifted system; singular-value exclusions, from the bound $\sigma_{\min}(\widehat{D}-s_i I) \le |\lambda - s_i|$, guarantee each interval $(s_i - \sigma_{i,\min},\, s_i + \sigma_{i,\min})$ is eigenvalue-free, so overlapping intervals from probe points $s_1 < s_2 < s_3$ certify a contiguous eigenvalue-free band whose right endpoint becomes the practical $X_L$; and the $\gamma_5$-Hermiticity relation recovers the left eigensystem from the right, so only the right projected subspace is needed.
Extended reading notes
Core claim
The central claim is that the signature of the Wilson–Dirac fermion determinant can be read reliably from a spectrally localized contour integral. Because $\gamma_5$-Hermiticity makes $\det(D)$ real and forces its sign to be set by $N_-$, the number of eigenvalues on the negative real axis, the problem reduces to counting those eigenvalues of the even Schur complement $\widehat{D}$. The contour integral is a Riesz spectral projector evaluated by finite quadrature, which returns the enclosed modes directly; the left contour boundary $X_L$ is fixed not by the Hermitian lower bound $\vartheta_{\min}(\widehat{H})$, which the authors find over-conservative, but by a sequence of singular values of shifted systems $\widehat{D} - s_i I$ certifying a contiguous eigenvalue-free interval. The authors state they have tested and verified that the procedure reproduces all negative-real solutions of $\widehat{D}$ up to machine precision, and hence the determinant signature, on a diverse subset of configurations from CLS ensembles, and that all negative-real eigenvalues are already resolved by the second FEAST iteration in every configuration tested.
Load-bearing premise
The whole counting rests on the singular-value exclusion producing a left contour boundary $X_L$ that is a true lower bound on the real parts of all eigenvalues: if a negative-real eigenvalue lies to the left of $X_L$, the contour never sees it and the sign comes out wrong.
Editorial extensions
If this is right
- For the configurations tested, the determinant sign is obtained in two FEAST iterations with eight quadrature nodes and a shifted-solver tolerance of $10^{-3}$, with contour-tuning overhead below a few percent of the total evaluation cost.
- The method resolves negative-real eigenvalues that differ by orders of magnitude, as in configuration C102r005n56 (solutions near $-1.74$ and shallow), at the price of $N_q = 12$ nodes and a tighter $10^{-5}$ tolerance.
- Because the singular-value machinery certifies a spectral boundary independently of explicitly returned eigenvalues, it can flag exceptional configurations whose low modes fall outside an assumed spectral range.
- The approach bypasses the spectral-flow requirement of following individual modes through a mass scan, giving direct access to the spectrum at the target quark mass.
- The independently solvable shifted systems at each quadrature node are well suited to straightforward parallelization, so the wall-time cost tracks the lattice half-volume and solver iterations.
Reading between the lines
- The singular-value exclusion procedure could be adapted to certify the lower spectral bound assumed by rational hybrid Monte Carlo (RHMC) approximations, identifying exceptional configurations whose low modes fall outside the assumed range and need dedicated reweighting — a direction the authors themselves note but do not develop.
- The same region-targeted contour filter applies outside lattice QCD, since isolating physically relevant eigenvalues embedded in dense complex spectra is a generic problem in non-Hermitian stability analyses and large-scale eigenvalue problems with expensive shifted solves.
- A natural stress test the paper does not run is a configuration engineered so a negative-real mode sits just to the left of the certified $X_L$, probing whether the hand-chosen buffers (0.02, 0.05, 0.15) and second-decimal rounding can be deceived.
Original abstract
Key properties of a large sparse operator can often be determined by only a small, spectrally localized subset of its eigenmodes. In lattice QCD with Wilson fermions, the sign of the fermion determinant is determined by the parity of the number of eigenvalues lying on the negative real axis of the non-Hermitian Wilson-Dirac operator. Determination of this signature can be challenging, particularly in parameter regimes where near-zero modes can lead to exceptional configurations. We present a contour-integral-based spectral projection method to count the negative-real eigenvalues of the non-Hermitian Wilson-Dirac operator $D$ that determine the sign of its determinant. The spectral projection is supplemented by an adapted deflation-accelerated solver for shifted linear systems and singular-value exclusions to reliably set spectral bounds and contour boundaries. We demonstrate the robustness of the method through systematic convergence tests, including variations of the contour size, and show its ability to resolve eigenvalues close to the negative real axis. Beyond fermion determinant sign evaluation, the approach provides a systematic means of isolating physically relevant eigenvalues embedded in dense complex spectra, with potential applications to non-Hermitian quantum systems, stability analyses, and large-scale eigenvalue problems.
The compiled record
The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
The abstract, sentence by sentence
Key properties of a large sparse operator can often be determined by only a small, spectrally localized subset of its eigenmodes. In lattice QCD with Wilson fermions, the sign of the fermion determinant is determined by the parity of the number of eigenvalues lying on the negative real axis of the non-Hermitian Wilson-Dirac operator. Determination of this signature can be challenging, particularly in parameter regimes where near-zero modes can lead to exceptional configurations. We present a contour-integral-based spectral projection method to count the negative-real eigenvalues of the non-Hermitian Wilson-Dirac operator $D$ that determine the sign of its determinant. The spectral projection is supplemented by an adapted deflation-accelerated solver for shifted linear systems and singular-value exclusions to reliably set spectral bounds and contour boundaries. We demonstrate the robustness of the method through systematic convergence tests, including variations of the contour size, and show its ability to resolve eigenvalues close to the negative real axis. Beyond fermion determinant sign evaluation, the approach provides a systematic means of isolating physically relevant eigenvalues embedded in dense complex spectra, with potential applications to non-Hermitian quantum systems, stability analyses, and large-scale eigenvalue problems.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | Key properties of a large sparse operator can often be determined by only a small, spectrally localized subset of its eigenmodes. | background 1.00 | 0.07 | 0.04 | hedged possibility 0.81 |
| s1 | In lattice QCD with Wilson fermions, the sign of the fermion determinant is determined by the parity of the number of eigenvalues lying on the negative real axis of the non-Hermitian Wilson-Dirac operator. | background 1.00 | 0.06 | 0.07 | asserted 1.00 |
| s2 | Determination of this signature can be challenging, particularly in parameter regimes where near-zero modes can lead to exceptional configurations. | background 0.72 | 0.05 | 0.02 | asserted 0.91 |
| s3 | We present a contour-integral-based spectral projection method to count the negative-real eigenvalues of the non-Hermitian Wilson-Dirac operator $D$ that determine the sign of its determinant. | contribution 0.69 | 0.95 | 0.49 | asserted 1.00 |
| s4 | The spectral projection is supplemented by an adapted deflation-accelerated solver for shifted linear systems and singular-value exclusions to reliably set spectral bounds and contour boundaries. | method 0.99 | 0.89 | 0.07 | asserted 1.00 |
| s5 | We demonstrate the robustness of the method through systematic convergence tests, including variations of the contour size, and show its ability to resolve eigenvalues close to the negative real axis. | result 0.84 | 0.96 | 0.19 | asserted 1.00 |
| s6 | Beyond fermion determinant sign evaluation, the approach provides a systematic means of isolating physically relevant eigenvalues embedded in dense complex spectra, with potential applications to non-Hermitian quantum systems, stability analyses, and large-scale eigenvalue problems. | interpretation 0.70 | 0.56 | 0.28 | asserted 0.55 |
What was examined, and what was not
- arXiv listing abstract v1
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
Figures
Figures reproduced from arXiv:2609.28646 by the authors.
References
Discussion
Disagree with the review, or know something it missed? Leave a comment, highlight any passage above to comment on it, or challenge this review formally.