Pith. sign in

REVIEW 6 minor 1 cited by

Localisation of pseudospectra on discrete groups

T0 review · 0 major / 6 minor · reviewed 2026-08-03 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper proves that pseudospectra of infinite band operators on ℓ²(G,Y) can be enclosed from finite local patches with explicit, vanishing error.

desk verdict Solid generalization of the authors' τ/τ1 enclosure methods to band operators on countable Abelian groups; the τ-method convergence claim needs qualification because the no-spectral-pollution hypothesis fails for the bilateral shift. read the letter →

arxiv 2607.29354 v1 pith:76NAIAET submitted 2026-07-31 math.SP math.FA

Simon N. Chandler-Wilde , Marko Lindner , Christian Seifert This is my paper · ORCID
classification math.SPmath.FA MSC 47A1047B3646E4047B80
keywords bandoperatorpseudospectrumlocalpatchestruncationpenaltyWieneralgebracountableAbeliangroupDirichletLaplacianspectralinclusion
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper generalizes two 'local patch' methods — the τ1 and τ truncation methods — from tridiagonal operators on ℓ²(Z) to band operators on ℓ²(G,Y) for any countable Abelian group G and Hilbert space Y. It proves that the pseudospectrum of such an operator is sandwiched between sets computed from finite submatrices ('local patches'), with a computable truncation penalty ε_p(W,A) that bounds the error. Under a subexponential-growth condition on a graph built from the operator's diagonals, this penalty tends to zero as the patch window W grows, so the enclosures converge to the true pseudospectrum in Hausdorff distance. This matters because spectra and pseudospectra of infinite matrices are central in physics and engineering, and these results turn them into finite, rigorously verified computations.

What carries the argument

The central mechanism is the truncation penalty ε(w,A) in the local inequality ||A_k x_k||/||x_k|| ≤ ||Ax||/||x|| + ε(w,A), arising from commutators of A with weighted truncation operators. The paper minimizes ε(w,A) over weight functions supported in the window W, expressing the optimal penalty as ε_p(W,A) = (Σ_j (||b^{(j)}||_∞ + ||b^{(-j)}||_∞)^{2-p})^{1/2} · sqrt(min Spec L_W^{(D)}), where L_W^{(D)} is the Dirichlet Laplacian on W for a graph (b,c) whose edge weights are p-th powers of symmetrized diagonal norms of A. Thus the whole error analysis reduces to the ground-state energy of a finite Laplacian, and its decay as W grows is governed by the graph's growth rate.

What would settle it

Take a band operator on ℓ²(Z) whose diagonal norms make the graph (b,c) a binary tree (e.g., b(2n,2n+1)=1) and compute ε_p(W_n,A) for growing intervals W_n; if ε_p(W_n,A) does not converge to 0, the universal convergence claim collapses. Alternatively, exhibit an A in W_p with such a graph and show the τ1 enclosures fail to Hausdorff-converge to Spec_ε A.

Watch

Extended reading notes

Core claim

For A belonging to the Wiener-type class W_p(E) on ℓ²(G,Y), the paper establishes two families of enclosures. The τ1 method gives a two-sided sandwich: Γ_ε^W(A) ⊂ Spec_ε A ⊂ Γ_{ε+ε_p(W,A)}^W(A), where the lower and upper sets are unions of pseudospectra of one-sided patch operators and their adjoints. The τ method gives a one-sided inclusion Spec_ε A ⊂ clos ∪_{k} Spec_{ε+ε_p(W,A)} A_k with square finite patches. When a sequence of windows exhausts G and the graph (b,c) formed from the diagonal norms has subexponential growth and is connected, the truncation penalty ε_p(W,A) converges to zero, so the τ1 enclosures converge to Spec_ε A from both sides and the τ enclosures converge provided the

Load-bearing premise

The convergence theorems assume the graph (b,c) built from the operator's diagonals has subexponential growth and is connected; if the graph grows faster, the truncation penalty need not vanish and the enclosures need not converge.

Editorial extensions

If this is right

  • For G=Z^d and band operators in W_0, the penalty ε_0(W_n,A) decays like 1/n, so the enclosures of Corollaries 5.10 and 5.17 converge at a known rate.
  • The τ1 method yields both an upper and a lower enclosure, so one can verify the absence of pseudospectrum in a region by checking finite patches alone, with a rigorous margin.
  • The τ method, despite requiring the no-spectral-pollution hypothesis (5.29), delivers a computable upper enclosure that converges to the pseudospectrum when that hypothesis holds; this includes many practically relevant operators.
  • The results cover all countable Abelian groups, including finitely generated groups Z^d ⊕ H via the tensor-product identification with ℓ²(Z^d, ℓ²(H)⊗Y), so cyclic and mixed lattice structures are handled uniformly.
  • Because ε_p(W,A) comes with an explicit formula in terms of a finite Dirichlet Laplacian, the error bound is computable in practice, not just asymptotic.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The subexponential-growth hypothesis may be necessary for convergence; if the diagonal graph is a regular tree (exponential growth), the penalty ε_p(W,A) likely fails to vanish, so the method would not converge — a testable boundary case.
  • The optimization over p ∈ [0,1] shows that the best error bound can occur at an interior point; a practical implementation could search over p for a given operator to tighten enclosures.
  • The connection to quasicrystal spectral gaps (via related local-patch work) suggests these enclosures can be used to certify spectral gaps in nonperiodic models on discrete groups, where the group is not Z^d but still countable Abelian.
  • The no-spectral-pollution condition (5.29) for the τ method might be characterized through limit operators or Fredholm theory; if a verifiable criterion is found, the τ method becomes fully automatic.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

0 major / 6 minor

Summary. The paper generalises the τ1 and τ truncation methods for computing spectral inclusion sets from tridiagonal operators on ℓ²(ℤ) to band and Wiener-class operators A on ℓ²(G,Y), where G is a countable Abelian group and Y a Hilbert space. The main technical results are quantitative commutator estimates: for every nonzero x there is a shifted patch with ‖A_k x̃_k‖/‖x̃_k‖ ≤ ‖Ax‖/‖x‖ + ε(w,A). Optimising the truncation weight w supported on a finite window W yields an explicit window-dependent penalty ε_p(W,A) expressed in terms of the ground state of a finite Dirichlet graph Laplacian (Propositions 5.3 and 5.14). This yields two-sided τ1 inclusions (Corollary 5.10) and one-sided τ inclusions (Corollary 5.17) for pseudospectra. Under subexponential growth and connectedness of the associated graph the τ1 enclosures Hausdorff-converge to the pseudospectrum (Corollary 5.11); the τ enclosures converge only under an extra no-spectral-pollution hypothesis (5.29) (Corollary 5.18). Examples discuss G=ℤ^d, finitely generated groups, and the optimal choice of p.

Significance. If correct, this is a valuable and clean extension of the authors' earlier work. The new bounds are explicit and directly computable from the diagonals and the window; the reduction of the weight optimisation to the principal eigenvalue of a Dirichlet graph Laplacian is elegant and makes the truncation penalty concrete. The τ1 results give genuine two-sided enclosures with a quantified, vanishing error, and the τ upper inclusions are useful for exclusion regions. The paper is transparent about its hypotheses, in particular the subexponential-growth condition and the (5.29) condition for τ convergence. The main caveat is that the advertised convergence for the τ method is conditional on an uncharacterised hypothesis that can fail (e.g., for the bilateral shift), which should be made explicit in the abstract and introduction.

minor comments (6)
  1. [Abstract, §1, Cor 5.18] The abstract and §1 state that the methods 'prove convergence to the spectrum, resp. pseudospectrum' without qualification. Cor 5.18 is conditional on (5.29). This condition is not characterized and fails for the bilateral shift A=V_1 on ℓ²(ℤ): each τ patch is an n×n nilpotent Jordan block, so 0∈limsup_n ∪_k Spec_ε A_{k,n} for every ε>0, while 0∉Spec_ε A for ε<1 because dist(0,σ(A))=1. Please qualify the abstract/introduction and add a remark; the upper inclusions (5.25)-(5.26) and τ1 results are unaffected.
  2. [Cor 5.18, Prop 5.14] Cor 5.18 refers to 'the graph (b^G,c^G) over G from Proposition 5.14', but Prop 5.14 defines only b^W and c^W for a finite W. State b^G and c^G explicitly (the proof uses b^G(k,ℓ)=2((‖b^{(k-ℓ)}‖+‖b^{(ℓ-k)}‖)/2)^p for k≠ℓ and c^G=0).
  3. [§5.2.3] After (5.23) the text says 'by the following elementary lemma:' and then gives a proof, but the lemma statement is missing. Please insert the lemma.
  4. [Cor 5.9, 5.10, 5.16, 5.17] The phrase 'A∈ W_p(E) as in (2.1)' is inaccurate; (2.1) defines band operators with finite J, while W_p includes infinite Wiener sums represented by (2.2). Change 'as in (2.1)' to 'as in (2.2)' throughout.
  5. [Prop 5.1 proof] The proof uses the summation index J in (5.7)-(5.10), but J is not defined in Prop 5.1; it should be G.
  6. [Throughout] Typos: 'subsexponential' in Cor 5.11 and Cor 5.18 should be 'subexponential'; 'developping' in §1; 'stetched' in Remark 5.2; 'choie' in §6.4.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the central inequalities are derived from scratch; the only caveat is that τ-method convergence is conditional on an uncharacterized no-spectral-pollution hypothesis.

full rationale

The central derivation chain (Prop. 5.1/5.13 → Prop. 5.3/5.14 → Cor. 5.9/5.16 → Cor. 5.10/5.17) is self-contained: the truncation bounds are obtained by commutator estimates, Minkowski's inequality, and Rayleigh quotients of Dirichlet Laplacians, none of which presupposes the spectrum/pseudospectrum inclusions being proved. The convergence of the τ1 enclosures (Cor. 5.11) uses external graph-Laplacian results [23] and standard resolvent-convergence arguments. Self-citations to [7] appear mostly as background, comparison (Sec. 6.1), and auxiliary technical facts (e.g., (5.27), Cor. 5.10(b)); they are not load-bearing for the main inequalities. The one caveat is Cor. 5.18: the τ-method convergence is explicitly conditional on the uncharacterized no-spectral-pollution hypothesis (5.29), which is essentially the upper half of the desired convergence. This is an honest limitation rather than a disguised derivation, but it means the abstract's unconditional-sounding 'prove convergence' is too broad for the τ method. This is a correctness/scope issue, not circularity.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

No new physical or mathematical entities are postulated; local patches and graph-Laplacian penalties are derived constructions. The only tunable parameter is the exponent p, which is not fitted to data. The load-bearing assumptions are mathematical domain conditions (subexponential growth, no spectral pollution) that are stated explicitly.

free parameters (1)
  • p (Wiener-class exponent)
    The exponent p ∈ [0,1] selects the class W_p(E) and enters the penalty ε_p(W,A). The theorems hold for any such p, but the sharpness of the bound depends on p; Examples 6.1–6.2 show the optimal p can vary with the operator. It is a user-chosen methodological parameter, not fitted to external data.
assumptions (4)
  • standard math Graph Laplacian results from Keller–Lenz–Wojciechowski [23]: under subexponential growth, min Spec L = 0, and Dirichlet Laplacians converge to the full Laplacian in generalized strong resolvent sense.
    Invoked in Corollaries 5.11 and 5.18 to conclude ε_p(W_n,A) → 0 as the windows exhaust G.
  • standard math Hilbert spaces (and ℓ²(G,Y)) have Globevnik's property, so ε ↦ spec_ε A is Hausdorff-continuous and clos spec_ε A = Spec_ε A.
    Used throughout the convergence arguments, especially in the sandwich argument of Corollary 5.11; cited to [16,17,28,29] and [4].
  • domain assumption The graph (b,c) constructed from A's diagonals in Propositions 5.3/5.14 has subexponential growth and is connected.
    The convergence theorems (Corollaries 5.11 and 5.18) depend on this to force min Spec L_{W_n}^{(D)} → 0 and hence vanishing truncation penalty.
  • domain assumption For the τ method, the patches (A_{k,n}) do not suffer from spectral pollution, i.e. (5.29) holds.
    Explicitly assumed in Corollary 5.18; without it the τ upper enclosures need not converge to Spec_ε A. The paper does not provide practical criteria for verifying it.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Localisation of pseudospectra on discrete groups." pith.science (2026). https://pith.science/paper/76NAIAET

@misc{pith2026260729354,
  author       = {Pith},
  title        = {Pith review of: Localisation of pseudospectra on discrete groups},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/76NAIAET}},
  note         = {Machine review of arXiv:2607.29354}
}
abstract

In this paper we generalise two of the methods and corresponding results from our previous paper ``On spectral inclusion sets and computing the spectra and pseudospectra of bounded linear operators'' [J. Spectr. Theory 14 (2024), 719--804] from tridiagonal operators on $\ell^2(\Z)$ to band operators $A$ on $\ell^2(G,Y)$ with a countable Abelian group $G$ and a Hilbert space $Y$. Again, we cover the pseudospectra of $A$, with error-control, via a union of pseudospectra of finite and moderately sized ``local patches'' of $A$. While a major application is to understand the case $G=\Z^d$ that is immanent in many physical problems, our new approach to the so-called $\tau$ and $\tau_1$ methods immediately extends to countable Abelian groups $G$.

Figures

Figures reproduced from arXiv: 2607.29354 by the authors.

Figure 1.1
Figure 1.1. The finite matrices that arise in the so-called [PITH_FULL_IMAGE:figures/full_fig_p001_1_1.png] view at source ↗
Figure 3.1
Figure 3.1. The honeycomb lattice G from Example 3.1 with its three generators, f1, f2 and f3. Note that we draw each fi here in terms of its action z 7→ fi +z and not primarily as a point of G. Also shown are an arbitrary node g ∈ G and the three hexagons Hj that it is part of as well as an element h ∈ G and its inverse −h whose location seems unexpected from the perspective of the zero element o ∈ G. In this case, (3.1) holds… view at source ↗
Figure 3.2
Figure 3.2. On the left we see in blue the integer grid generated by ( [PITH_FULL_IMAGE:figures/full_fig_p008_3_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Spectral and Pseudospectral Approximation of Finite-Interaction-Range Operators in Doubling Metric Measure Spaces

    math.SP 2026-08 conditional novelty 6.0 of 10

    Window-local lower norms approximate the global lower norm of finite-interaction-range operators with explicit O(1/L) error on doubling metric measure spaces, yielding rigorous pseudospectral inclusions.

Reference graph

Works this paper leans on

33 extracted references · 1 canonical work pages · cited by 1 Pith paper

  1. [7]

    S. N. Chandler-Wilde, R. Chonchaiya and M. Lindner: On spectral inclusion sets and computing the spectra and pseudospectra of bounded linear operators', Journal of Spectral Theory 14 (2024), 719--804

    2024

  2. [1]

    Avron, P.H.M

    J. Avron, P.H.M. v.Mouche and B. Simon : On the measure of the spectrum for the almost Mathieu operator, Communications in Mathematical Physics 132 , 103--118 (1990)

    1990

  3. [2]

    Ben-Artzi, A

    J. Ben-Artzi, A. C. Hansen, O. Nevanlinna and M. Seidel : New barriers in complexity theory: On the Solvability Complexity Index and towers of algorithms, C. R. Acad. Sci. Paris, Ser. I 353 (2015), 931--936

    2015

  4. [3]

    Ben-Artzi, M

    J. Ben-Artzi, M. J. Colbrook, A. C. Hansen, O. Nevanlinna and M. Seidel : Can everything be computed? - On the Solvability Complexity Index and Towers of Algorithms, arXiv:1508.03280v5 https://arxiv.org/abs/1508.03280v5 (2020)

    arXiv 2020

  5. [4]

    Boyko and V

    N. Boyko and V. Kadets : Uniform G-convexity for vector-valued Lp spaces, Serdica Math. J. 35 (2009), 1--14

    2009

  6. [5]

    S. N. Chandler-Wilde, R. Chonchaiya and M. Lindner: On the Spectra and Pseudospectra of a Class of Non-Self-Adjoint Random Matrices and Operators, Operators and Matrices 7 , 739--775 (2013)

    2013

  7. [6]

    S. N. Chandler-Wilde, R. Chonchaiya and M. Lindner: Convergent spectral inclusion sets for banded matrices, Proc. Appl. Math. Mech. 23 (2023), e202300016. First published: 4 October 2023 https://doi.org/10.1002/pamm.202300016

    work page doi:10.1002/pamm.202300016 2023

  8. [8]

    S. N. Chandler-Wilde and M. Lindner: Something with semi-infinite matrices and essential spectra , in preparation

Show all 33 references
  1. [9]

    S. N. Chandler-Wilde and M. Lindner: Gershgorin-type spectral inclusions for matrices, Linear Algebra and its Applications 732 (2025), 33--73

    2025

  2. [10]

    Choi, G.A

    M.-D. Choi, G.A. Elliot and N. Yui : Gauss polynomials and the rotation algebra, Inventiones mathematicae 99 , 225--246 (1990)

    1990

  3. [11]

    Chonchaiya : Computing the Spectra and Pseudospectra of Non-Self-Adjoint Random Operators Arising in Mathematical Physics, PhD thesis , University of Reading, 2010

    R. Chonchaiya : Computing the Spectra and Pseudospectra of Non-Self-Adjoint Random Operators Arising in Mathematical Physics, PhD thesis , University of Reading, 2010

    2010

  4. [12]

    Cohn : Algebra , vol

    P. Cohn : Algebra , vol. I, Wiley, 1974

    1974

  5. [13]

    M. J. Colbrook, M. Embree and J. Fillman: Optimal Algorithms for Quantifying Spectral Size with Applications to Quasicrystals, arXiv preprint 2407.20353 (2024)

    2024 arXiv

  6. [14]

    M. J. Colbrook and A. C. Hansen: The foundations of spectral computations via the Solvability Complexity Index hierarchy, J. Eur. Math. Soc. 25 (2023), 4639--4718

    2023

  7. [15]

    M. J. Colbrook: Infinite-Dimensional Spectral Computations: Foundations, Algorithms, and Modern Applications , Cambridge University Press, 2024 (in preparation)

    2024

  8. [16]

    Globevnik and I

    J. Globevnik and I. Vidav : On operator-valued analytic functions with constant norm, J. Funct. Anal. 15 (1974), 394–403

    1974

  9. [17]

    Globevnik : Norm-constant analytic functions and equivalent norms, Illinois J

    J. Globevnik : Norm-constant analytic functions and equivalent norms, Illinois J. Math. 20 (1976), 503--506

    1976

  10. [18]

    Hagen, S

    R. Hagen, S. Roch and B. Silbermann: C^* -Algebras and Numerical Analysis , Marcel Dekker, Inc., New York, Basel, 2001

    2001

  11. [19]

    Hagger, M

    R. Hagger, M. Lindner , and M. Seidel: Essential pseudospectra and essential norms of band-dominated operators, J.\ Math.\ Anal.\ Appl.\@ 437 (2016), 255--291

    2016

  12. [20]

    Hege : Spectral gaps in systems of finite local complexity , PhD thesis , University of T\"ubingen, 2024

    P. Hege : Spectral gaps in systems of finite local complexity , PhD thesis , University of T\"ubingen, 2024

    2024

  13. [21]

    P. Hege, M. Moscolari and S. Teufel : Finding spectral gaps in quasicrystals, Phys. Rev. B 106 , 155140

  14. [22]

    P. Hege, M. Moscolari and S. Teufel : Computing the spectrum and pseudospectrum of infinite-volume operators from local patches, Math. Comp. 95 (358), 833--866 (2026)

    2026

  15. [23]

    Keller, D

    M. Keller, D. Lenz and R. Wojciechowski : Graphs and Discrete Dirichlet Spaces , Grundlehren der mathematischen Wissenschaften, 358, Springer, 2021

    2021

  16. [24]

    Lindner : Infinite Matrices and their Finite Sections: An Introduction to the Limit Operator Method , Frontiers in Mathematics, Birkh\"auser 2006

    M. Lindner : Infinite Matrices and their Finite Sections: An Introduction to the Limit Operator Method , Frontiers in Mathematics, Birkh\"auser 2006

    2006

  17. [25]

    Lindner and M

    M. Lindner and M. Seidel : An Affirmative Answer to a Core Issue on Limit Operators, Journal of Functional Analysis 267 , 901--917 (2014)

    2014

  18. [26]

    V. S. Rabinovich, S. Roch and B. Silbermann : Limit Operators and Their Applications in Operator Theory , Birkh\"auser 2004

    2004

  19. [27]

    Reed and B

    M. Reed and B. Simon: Methods of Modern Mathematical Physics, I. Functional Analysis , Academic Press, New York, San Francisco, London 1975

    1975

  20. [28]

    Shargorodsky : On the level sets of the resolvent norm of a linear operator, Bulletin of the LMS 40 (2008), 493--504

    E. Shargorodsky : On the level sets of the resolvent norm of a linear operator, Bulletin of the LMS 40 (2008), 493--504

    2008

  21. [29]

    Shargorodsky: On the definition of pseudospectra, Bull

    E. Shargorodsky: On the definition of pseudospectra, Bull. London Math. Soc. 41 (2009), 524--534

    2009

  22. [30]

    L. N. Trefethen and M. Embree : Spectra and Pseudospectra: The Behavior of Nonnormal Matrices and Operators , Princeton University Press, Princeton, NJ, 2005

    2005

  23. [31]

    Wittig : Spectral Enclosures for Non-Self-Adjoint Operators , Master Thesis, Uni Hamburg & TU Hamburg, 2026

    M. Wittig : Spectral Enclosures for Non-Self-Adjoint Operators , Master Thesis, Uni Hamburg & TU Hamburg, 2026

    2026

  24. [32]

    Wittig : Spectral and Pseudospectral Approximation of Finite-Interaction-Range Operators in Doubling Metric Measure Spaces, in preparation

    M. Wittig : Spectral and Pseudospectral Approximation of Finite-Interaction-Range Operators in Doubling Metric Measure Spaces, in preparation

  25. [33]

    T. G. Wright and L. N. Trefethen : Pseudospectra of rectangular matrices, IMA Journal of Numerical Analysis 22 (2002), 501--519

    2002

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.