Pith. sign in

REVIEW 4 major objections 3 minor 19 references

Generalized Multiple Operator Integrals and Perturbation Theory for Operators with Continuous Spectra

T0 review · 4 major / 3 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read This paper claims that generalized multiple operator integrals can be defined for non-normal operators with continuous spectra, and that a Krein-type spectral shift formula follows.

desk verdict The central spectral decomposition (Eq. 2/18) fails for the unilateral shift, so the GMOI calculus is undefined on basic non-normal continuous-spectrum operators; the formal machinery doesn't overcome this. read the letter →

arxiv 2507.23049 v1 pith:PTLBLOJ2 submitted 2025-07-30 math.FA math.OA

Shih-Yu Chang This is my paper · ORCID
classification math.FAmath.OA MSC 47A5547A60
keywords multipleoperatorintegralsgeneralizedcontinuousspectrumnon-self-adjointoperatorsperturbationformulaspectralshiftfunctiontracedivideddifferences
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 sets out to build a generalized multiple operator integral (GMOI) calculus for operators that need not be self-adjoint or normal and may have continuous spectra. It proposes a definition in Eq. (17) that extends the usual multiple operator integral by incorporating nilpotent correction terms coming from a Jordan-type spectral decomposition of each operator. On top of that definition it proves a perturbation formula, norm and Lipschitz bounds, operator-norm continuity, and a Krein-type spectral shift representation for the trace of Taylor remainders. A sympathetic reading is that this supplies a unified multi-operator integration framework for a class of continuous-spectrum non-normal operators that classical spectral-theoretic tools cannot handle.

What carries the argument

The load-bearing object is the decomposition $X = \int \lambda\,dE_X(\lambda) + \int (X-\lambda I)\,dE_X(\lambda)$ used for every operator, with $dE_X(\lambda)$ treated as spectral measures and $m_\lambda$ as nilpotent orders so that $(X-\lambda I)^{m_\lambda}dE_X(\lambda)=0$. This decomposition is what converts non-commuting, non-normal operators into objects on which divided differences and Taylor expansions can be integrated. The paper combines this with the $\ell$-th divided difference $\beta^{[\ell]}$ as the integral symbol, a binary-index expansion of nilpotent parts, and Riesz-representation trace duality to obtain the spectral shift functions $\eta_{i,j}$ and kernels $K_{i',j'}$.

What would settle it

Take a bounded non-normal operator without eigenvalues, such as the Volterra operator $Vf(x)=\int_0^x f(t)\,dt$ on $L^2[0,1]$, and check whether the assumed decomposition $V = \int \lambda\,dE_V(\lambda)+\int (V-\lambda I)\,dE_V(\lambda)$ can be realized with nontrivial nilpotent orders; if the spectral measures $dE_V(\lambda)$ and orders $m_\lambda$ cannot be defined, then the right-hand side of Eq. (17) is undefined and the central claim fails for this operator.

Watch

Extended reading notes

Core claim

The central claim is that $T^{X_1,\ldots,X_{\zeta+1}}_{\beta}(Y_1,\ldots,Y_\zeta)$, defined by Eq. (17), is a well-behaved operator whenever each parameter and argument operator decomposes as $X = \int \lambda\,dE_X(\lambda) + \int (X-\lambda I)\,dE_X(\lambda)$ with nilpotent orders $m_\lambda$. With this definition, Theorems 2, 5, and 8 provide a perturbation formula, continuity under operator-norm convergence, norm and Lipschitz estimates, and an $(n-1)$-th order spectral shift formula. The paper also shows that the conventional multiple operator integral is a special case of the spectral mapping picture, so the framework extends rather than replaces classical theory. In the spectral shift application, the trace of an $n$-th order Taylor remainder is shown to split into single-variable integrals weighted by functions $\eta_{i,j}(z)$ and double integrals with bivariate Hilbert–Schmidt kernels $K_{i',j'}(z_1,z_2)$.

Load-bearing premise

The paper assumes every operator decomposes as $X = \int \lambda\,dE_X(\lambda) + \int (X-\lambda I)\,dE_X(\lambda)$ with nilpotent orders $m_\lambda$, a Jordan-like spectral structure that is known to hold in finite dimensions or for spectral operators, but no theorem in the paper or its cited sources establishes this decomposition for general non-normal operators with continuous spectra.

Editorial extensions

If this is right

  • The GMOI norm and Lipschitz bounds give explicit control over how the integral changes when the argument operators $Y_i$ are perturbed, extending classical multiple operator integral estimates to continuous-spectrum non-normal settings.
  • The perturbation formula of Theorem 2 yields a telescoping argument that converts operator-norm convergence $X_{j,\ell_j} \to X_j$ into convergence of the corresponding GMOIs, provided the divided difference function is sufficiently regular.
  • The first-order Krein spectral shift formula $\mathrm{Tr}(f(X+Y)-f(X)) = \int f^{(1)}(z)\eta_1(z)\,dz$ is established for generalized double operator integrals with continuous spectra.
  • For Taylor remainders of arbitrary order, the trace admits an integral representation with single-variable functions $\eta_{i,j}$ and bivariate Hilbert–Schmidt kernels $K_{i',j'}$, giving a higher-order spectral shift structure.
  • Because the conventional MOI is recovered as a special case, the new formalism is consistent with classical multiple operator integration when the operators are normal or self-adjoint.

Reading between the lines

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

  • Editorial inference: if the assumed spectral decomposition can be justified for a broader class of operators, such as spectral operators or operators with a Riesz–Dunford functional calculus, the same formulas would give trace and scattering identities for concrete continuous-spectrum examples in mathematical physics.
  • Editorial inference: the bivariate kernels $K_{i',j'}$ likely encode cross-talk between different derivative orders of $f$, and one could test whether they reduce to derivatives of the first-order spectral shift function in special cases such as self-adjoint $X$ and $Y$.
  • Editorial inference: the Lipschitz constant $\Upsilon_i$ from Theorem 4 could be estimated numerically on finite-rank truncations of continuous-spectrum operators; a divergence as the truncation dimension grows would localize exactly which hypothesis of the framework fails.
  • Editorial inference: the paper's formalism suggests a hierarchy of higher-order spectral shift functions, but whether these functions are unique or satisfy consistency relations analogous to Krein's classical uniqueness is not established in the text.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 3 minor

Summary. The paper proposes a theory of generalized multiple operator integrals (GMOIs) for general non-normal, non-self-adjoint operators with continuous spectra. The central definition in Eq. (17) is based on a Jordan-type spectral decomposition Eq. (2)/(18), and the paper claims a perturbation formula (Theorem 2), norm and Lipschitz estimates (Theorems 3 and 4), operator-norm continuity (Theorem 5), and a Krein-type spectral shift formula for GDOIs and its (n−1)-th order approximation (Theorems 6 and 8).

Significance. If the claimed framework were valid, it would substantially extend multilinear operator integration beyond the self-adjoint and normal settings and would give new spectral shift representations. The paper is explicit and ambitious: it defines the objects concretely, derives detailed algebraic expressions in Eqs. (37) through (47), and attempts to prove two key perturbation identities. I also note that no machine-checked proofs or reproducible code are provided, and the central theorems rely on the author's own unpublished preprints [15]–[18]. However, the load-bearing spectral decomposition in Eq. (2)/(18) is not valid for the claimed operator class and fails for the unilateral shift, so the significance of the results as stated is not realized.

major comments (4)
  1. [Section 2, Eqs. (2) and (18); Section 3.1, Eq. (17)] The paper assumes every bounded non-normal operator with continuous spectra decomposes as X = ∫ λ dE_X(λ) + ∫ (X − λI) dE_X(λ), where dE_X is a spectral measure and m_λ is a nilpotent order. This is a Jordan-type decomposition that exists for finite-dimensional operators and for spectral operators in Dunford's sense, but it is not available for general non-normal continuous-spectrum operators. A concrete counterexample is the unilateral shift S on ℓ², which is bounded and non-normal, has spectrum the closed unit disk, and is not a spectral operator; it admits no countably additive projection-valued resolution of the identity. Therefore Eq. (17), which defines every GMOI from this decomposition, is undefined for such a basic example. The citation of Theorem 1 from [18] does not remedy this, because Theorem 1 itself assumes the same decomposition in Eq. (5).
  2. [Section 4, Theorem 2, proof of Eq. (49), Eq. (57)] The proof of the first perturbation identity uses ∫ dE_C(λ_c) = I = ∫ dE_D(λ_d), which presupposes that dE_C and dE_D are projection-valued resolutions of the identity. This is exactly the spectrality assumption that is not established for the operator class in question. In addition, only the identities in Eqs. (49) and (50) are proved; the identities in Eqs. (51) through (56) are dismissed with 'can be proved similarly' even though they are needed for the theorem.
  3. [Section 6, Lemma 3 and Theorem 5] Lemma 3's proof relies on 'smooth variation of the Jordan decomposition (i.e., assuming stable geometric multiplicities)' and on Fréchet derivatives of spectral projectors and nilpotent parts. For non-spectral operators with continuous spectra these objects are not defined, and the asserted convergence of the correction terms to zero is not established. Since Theorem 5 applies Theorem 2 and Lemma 3 to every term of the telescoping sum in Eq. (98), the continuity claim is unsupported.
  4. [Section 7, Lemma 4, Theorem 8, and footnote 1] Theorem 8 asserts the existence of single-variable functions η_{i,j}(z) and bivariate kernels K_{i',j'}(z1,z2) through the Riesz representation theorem and a 'generalized Riesz representation theorem.' No hypotheses are provided under which the linear functional f^(i) ↦ Tr(T_{f^{[i]}}) and the bilinear functionals in Lemma 4 are bounded or representable; moreover, for operators outside the scope of Eq. (18) the GMOIs in Eq. (113) are undefined. The proof also depends on Theorem 7, which is quoted from [17], and footnote 1 explicitly states that the continuous-spectrum extension 'preserves the same structural form' as Theorem 2; this is an asserted analogy, not a proof, and it does not supply the missing justification.
minor comments (3)
  1. [Throughout] The manuscript contains numerous typographical errors and misspellings, such as 'comventional,' 'signficant,' 'beings complicaiton,' 'terems,' and 'funtion,' together with repeated broken equations, for example the double equality sign in Eq. (7) and the missing parenthesis in Eq. (45) ('dED(λdYj'). A careful proofreading pass is needed.
  2. [Section 2, Eqs. (8), (10), and (13)] The notation is inconsistent in several places: 'λ kζ+1' and 'λ k′' appear instead of the intended eigenvalue labels, and the relationship between the indices in Eq. (10) and the arguments of β is not fully explained.
  3. [Section 3.1, Eq. (17)] The definition introduces Y_{ζ+1} = I at the end of the formula, but this convention is not stated before the integral is written; it should be made explicit earlier, since it affects the product structure in every term of the definition.

Circularity Check

4 steps flagged · score 8.0 of 10

The GMOI construction rests on an unproved Jordan-type spectral decomposition imported from the author's own preprints, and the perturbation and Krein-type spectral shift results are delegated to self-citations or Riesz-representation restatements.

  1. ansatz smuggled in via citation [Section 2, Eq. (2) and Section 3.1, Eq. (17); Theorem 1 from [18]]
    "According to Theorem 1, GMOIs can be defined as follows. ... Xp = ∫ λp dEXp(λp) + ∫ (Xp − λp I)dEXp(λp) (18). Theorem 1: Given ... and the operator Xl decomposed by: Xl = ∫ λl dEXl(λl) + ∫ (Xl − λl I)dEXl(λl) (5)."

    The GMOI definition is presented as a consequence of Theorem 1, but Theorem 1, cited from the author's arXiv:2411.11883, has as its hypothesis exactly the same decomposition (5) that is restated as Eq. (18). No theorem in the paper or in the cited source proves that a general non-normal continuous-spectrum operator admits such a Jordan-type resolution with projection-valued measures and nilpotent orders; the unilateral shift is a standard counterexample. The continuous-spectrum GMOI therefore reduces to the same unverified ansatz imported from a same-author citation.

  2. self citation load bearing [Section 4, Lemma 1 and Theorem 2 proof]
    "Lemma 1 (adopted from Lemma 2 in [17]) ... Proof: The proof of this Lemma can be found at Lemma 2 in [17]. ... The proof idea follows the proof in Theorem 3 from [17]."

    Theorem 2, the paper's perturbation formula, is not proved in the continuous-spectrum setting; its proof is delegated to Theorem 3 in the author's finite-dimensional preprint [17], and Lemma 1's proof is likewise merely located in [17]. The paper states elsewhere that the continuous-spectrum Theorem 2 'preserves the same structural form' as the matrix result. Thus the central analytic step is carried by a load-bearing self-citation chain: the claimed generalization inherits its validity from unpublished same-author work rather than from an independent continuous-spectrum argument.

2 more flagged steps
  1. self definitional [Section 7, Theorem 6]
    "According to the Generalized Double Operator ntegral (DOI) framework (Theorem 4 in [16]) ... This expression defines a linear functional f→Tr(f(X+Y)−f(X)) on the space of derivatives f, and by the Riesz representation theorem, there exists a function η1(z) such that Tr(f(X+Y)−f(X)) = ∫ f(1)(z)η1(z)dz. This η1(z) is called the first-order spectral shift function."

    The advertised Krein-type spectral shift formula is obtained by applying Riesz representation to the trace functional of a DOI, whose definition already presupposes Eq. (2). Existence of η1 is then true by definition of the representing function; no property of a spectral shift function is derived, no boundedness of the functional is established, and the formula is exactly the abstract representation statement. The 'Krein-type' content is renamed, not proved.

  2. self definitional [Section 7, Theorem 8 and its proof]
    "From Lemma 4, as f(i) → Tr(T_{f[i]}^{Pi(j)}([Z1,...,Zi]j)) is a linear functional, from Riesz representation theorem, we can find functions η_{i,j}(z) to have ... there exist single-variable functions η_{i,j}(z) and bivariate functions K_{i′,j′}(z1,z2) such that Tr(Rn(f,X,Y)) = ... ."

    Theorem 8's conclusion, the existence of η_{i,j} and K_{i′,j′}, is the same statement as the Riesz-representation invocation in its proof: every linear functional on f^(i) has a representing function and every bilinear one has a kernel. No bound is proved and no kernel is computed; the (n−1)-th order spectral shift formula is therefore a restatement of the definition of the representing kernels of the trace functionals, not a derived spectral identity.

full rationale

Score 8 because the central construction is forced by the paper's own unchallenged premise, imported from same-author preprints. Eq. (2)/(18) postulates a Jordan-type spectral decomposition for all non-normal operators with continuous spectra; Eq. (17) defines GMOIs 'according to Theorem 1,' whose hypothesis is exactly (5). Theorem 2, Lemma 1, and Theorem 6/7 are delegated to the same author's preprints [17]/[16]/[18], and Theorem 8's spectral-shift existence is just Riesz representation. This is not a case of independent benchmark verification: no external theorem, machine-checked proof, or computation outside the self-cited chain supplies the missing decomposition. Separately, Lemma 3's 'smooth variation of the Jordan decomposition' assumption is an additional unsupported condition for the continuity theorem, so Theorem 5 is conditional on the very spectral structure the paper claims to handle. If the decomposition is granted, the subsequent algebra is internally coherent for spectral operators with nilpotent parts, so the circularity is not a literal equation-to-equation identity; rather, the advertised general continuous-spectrum claim reduces to an assumed premise plus a self-citation chain. This warrants 8 rather than 10.

Assumptions & free parameters 0 free parameters · 4 assumptions · 2 invented entities

The paper's central claim rests on a spectral decomposition (Eq. (18)) that is not valid for generic non-normal continuous-spectrum operators, on transfer of finite-dimensional results from the author's prior preprints [17,18], and on unstated trace-class assumptions in the spectral shift section. No free parameters are fitted.

assumptions (4)
  • ad hoc to paper Every operator X_p admits a decomposition X_p = ∫ λ dE_Xp(λ) + ∫ (X_p − λ I) dE_Xp(λ), with spectral measures dE_Xp(λ) and nilpotent orders m_λ.
    Invoked in Section 2 (Eq. (2)) and Section 3.1 (Eq. (18)) for general non-normal, non-self-adjoint operators with continuous spectra; no proof or reference establishes such a decomposition for these operators.
  • domain assumption Stable geometric multiplicities and smooth variation of the Jordan decomposition under perturbation C → D.
    Used in Lemma 3 proof to claim the correction terms are O(t); this is standard for finite-dimensional Jordan blocks but not for continuous-spectrum operators.
  • ad hoc to paper The trace of GMOIs is well-defined and finite, and Riesz representation applies to the functionals f ↦ Tr(T^{...}_{f^{[i]}}...).
    Theorem 6 and Theorem 8 assume trace-class behavior and an ambient Hilbert or L^2 space without stating conditions on X, Y, or the function space.
  • domain assumption All partial derivatives of β[ζ+1] are bounded.
    Assumed in Theorem 5 to obtain continuity; not established for the functions used in the Krein-type applications.
invented entities (2)
  • Spectral shift function η_1(z) for non-self-adjoint continuous-spectrum operators
    purpose: Represents the linear functional f ↦ Tr(f(X+Y)-f(X)) as ∫ f^(1)(z) η_1(z) dz in Theorem 6
    No construction is given; existence is asserted via Riesz representation without specifying the measure space or proving boundedness.
  • Bilinear kernels K_{i',j'}(z1,z2)
    purpose: Represents traces of the correction terms in Theorem 8 as ∫∫ f^{i'+1}(z1) K(z1,z2) f^{i'}(z2) dz1 dz2
    Asserted to exist by a 'generalized Riesz representation theorem', but the theorem is not stated or proved, and no explicit kernel is derived.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Generalized Multiple Operator Integrals and Perturbation Theory for Operators with Continuous Spectra." pith.science (2026). https://pith.science/paper/PTLBLOJ2

@misc{pith2026250723049,
  author       = {Pith},
  title        = {Pith review of: Generalized Multiple Operator Integrals and Perturbation Theory for Operators with Continuous Spectra},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PTLBLOJ2}},
  note         = {Machine review of arXiv:2507.23049}
}
read the original abstract

Operators with continuous spectra naturally arise in spectral theory, quantum mechanics, automorphic forms, and noncommutative geometry. However, analyzing such operators, particularly in the non-selfadjoint setting, remains challenging due to spectral instability and the lack of an orthonormal basis. This work advances the theory of Multiple Operator Integrals (MOIs) by developing a unified framework for generalized MOIs (GMOIs) associated with general (non-normal, non-selfadjoint) operators possessing continuous spectra. Building on prior work in Generalized Double Operator Integrals (GDOIs) and finite dimensional GMOIs, we extend the theory to include: the formulation of GMOIs in the continuous spectrum setting, their algebraic structure, continuity properties, norm and Lipschitz estimates, and a perturbation formula that generalizes classical results. As a key application, we derive a Krein-type spectral shift formula for GDOIs in the continuous spectrum setting and further extend it to arbitrary-order approximations. These contributions provide a foundation for broader developments in spectral theory, operator algebras, noncommutative geometry, and noncommutative analysis.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

19 extracted references · 18 canonical work pages

  1. [17]

    Generalized Multiple Operator Integrals for Operators with Finite Dimensions

    ——, “Generalized multiple operator integrals for operators with finite dimensions,”arXiv preprint arXiv:2506.19971, 2025

    work page Pith review arXiv 2025

  2. [18]

    Operator characterization via projectors and nilpotents,

    ——, “Operator characterization via projectors and nilpotents,”arXiv preprint arXiv:2411.11883, 2024

    arXiv 2024

  3. [15]

    Generalized Double Operator Integrals: Finite Dimensions

    S.-Y . Chang, “Generalized double operator integrals: Finite dimensions,”arXiv preprint arXiv:2503.16239, 2025

    work page Pith review arXiv 2025

  4. [1]

    Iwaniec,Topics in Classical Automorphic Forms

    H. Iwaniec,Topics in Classical Automorphic Forms. American Mathematical Society, 1997

    work page 1997

  5. [2]

    D. A. Hejhal,The Selberg Trace Formula for PSL(2,R), Vol. I. Springer, 1976

    work page 1976

  6. [3]

    Sarnak,Some Applications of Modular Forms

    P. Sarnak,Some Applications of Modular Forms. Cambridge University Press, 1990

    work page 1990

  7. [4]

    Trace formula in noncommutative geometry and the zeros of the riemann zeta function,

    A. Connes, “Trace formula in noncommutative geometry and the zeros of the riemann zeta function,” Selecta Mathematica, vol. 5, no. 1, pp. 29–106, 1999

    work page 1999

  8. [5]

    A probabilistic approach to the leader problem in random graphs

    A. Connes and C. Consani, “The riemann–roch strategy: spectral realization of zeros of l-functions,” https://arxiv.org/abs/1703.09908, 2017, arXiv:1703.09908

    work page Pith review arXiv 2017

Show all 19 references
  1. [6]

    Connes,Noncommutative geometry

    A. Connes,Noncommutative geometry. Springer, 1994

    1994

  2. [7]

    R. P. Langlands,On the Functional Equations Satisfied by Eisenstein Series. Springer, 1976

    1976

  3. [8]

    An introduction to the trace formula,

    J. Arthur, “An introduction to the trace formula,” inHarmonic Analysis, the Trace Formula, and Shimura Varieties. American Mathematical Society, 2005, pp. 1–263

    2005

  4. [9]

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

    2005

  5. [10]

    E. B. Davies,Linear Operators and their Spectra. Cambridge University Press, 2007

    2007

  6. [11]

    Real spectra in non-hermitian hamiltonians havingPTsymmetry,

    C. M. Bender and S. Boettcher, “Real spectra in non-hermitian hamiltonians havingPTsymmetry,” Physical Review Letters, vol. 80, no. 24, pp. 5243–5246, 1998

    1998

  7. [12]

    Making sense of non-hermitian hamiltonians,

    C. M. Bender, “Making sense of non-hermitian hamiltonians,”Reports on Progress in Physics, vol. 70, no. 6, pp. 947–1018, 2007

    2007

  8. [13]

    Engel and R

    K.-J. Engel and R. Nagel,One-Parameter Semigroups for Linear Evolution Equations. Springer, 2000

    2000

  9. [14]

    Zworski,Mathematical Theory of Scattering Resonances

    M. Zworski,Mathematical Theory of Scattering Resonances. American Mathematical Society, 2017. 46

    2017

  10. [16]

    Generalized double operator integrals for continuous spectrum operators,

    ——, “Generalized double operator integrals for continuous spectrum operators,”arXiv preprint arXiv:2505.02188, 2025

    2025 arXiv

  11. [19]

    Skripka and A

    A. Skripka and A. Tomskova,Multilinear operator integrals theory and applications. Springer, 2019. 47

    2019

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.