Pith. sign in

REVIEW 5 minor 31 references

A fractal analogue of Lieb's Hardy–Littlewood–Sobolev inequality for orthonormal functions holds for Frostman measures.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review

2026-08-01 22:13 UTC pith:R6GN6FSU

load-bearing objection Genuinely new fractal orthonormal Sobolev inequalities; the proof is essentially correct, and the one suspicious interpolation step in §5 checks out.

arxiv 2607.15826 v1 pith:R6GN6FSU submitted 2026-07-17 math.CA math-phmath.MPmath.SP

Orthonormal Sobolev estimates with fractal measures

Neal Bez , Keith M. Rogers , Shunya Toyoshima This is my paper
classification math.CA math-phmath.MPmath.SP MSC 42B2046E3547B1081Q10
keywords orthonormal functionsHardy–Littlewood–Sobolev inequalityFrostman measurestrace inequalitieseigenvalue boundsshell potentialsanalytic interpolationfractal measures
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper establishes a fractal analogue of the classical Hardy–Littlewood–Sobolev inequality for orthonormal functions: for any α-Frostman measure μ, the L^{q/2}(dμ)-norm of a sum of squared fractional derivatives of an orthonormal system is controlled by the ℓ^{p/2}-norm of the coefficients, for 2 ≤ p < q < ∞ and s = d/2 − α/q. This generalizes Lieb's bound for orthonormal systems and can be viewed as a trace theorem for orthonormal functions on lower-dimensional sets. The inequality also yields an interaction energy bound for two distinct Frostman measures, and, as applications, the authors recover known bounds for the number and sum of negative eigenvalues of Schrödinger operators with shell potentials. The proof is direct, using Fourier analysis, a reproof of Adams' fractal Hardy–Littlewood–Sobolev inequality, and analytic interpolation.

Core claim

The central claim is Theorem 1.1: for 2 ≤ p < q < ∞ and s = d/2 − α/q, there is a constant C such that ‖∑_j γ_j |D^{-s} f_j|²‖_{L^{q/2}(dμ)} ≤ C [μ]_α^{2/q} ‖γ‖_{ℓ^{p/2}} whenever (f_j) is orthonormal in L²(R^d) and μ is an α-Frostman measure. This is a direct generalization of Lieb's Hardy–Littlewood–Sobolev inequality for orthonormal systems to the fractal setting. From it the authors derive a more general trace inequality (Theorem 1.2) for orthonormal functions on lower-dimensional sets, and an interaction energy estimate (Corollary 1.2) for two Frostman measures with distinct dimensions. In the final section, these estimates are shown to recover the Rozenblum–Tashchiyan bound for the num

What carries the argument

The key mechanism is an analytic interpolation between two endpoint estimates. The authors define an analytic family of operators T_z mapping a function g to the sequence (g D^z f_j)_j, where D^z is the fractional derivative of complex order z. On a strip in the complex plane, they interpolate between a weak-type L^{q/2,∞} estimate (Theorem 3.1) and a strong-type estimate derived from a Fourier-analytic reproof of Adams' fractal Hardy–Littlewood–Sobolev inequality (Theorem 4.3). A real interpolation trick that separates high and low frequencies is used to obtain the endpoint estimates. The dual operator (D^{-s})^* plays a role by swapping the roles of Lebesgue measure and the Frostman measur

Load-bearing premise

The proof's central step is an analytic interpolation on a strip of complex exponents, where the authors assert a uniform bound for the operator family T_z; this step is not fully detailed in the paper.

What would settle it

Evaluate the operator norm of T_z for a fixed orthonormal system (e.g., Hermite functions) and an explicit Frostman measure (e.g., a Cantor measure) as z approaches the boundary of the strip in the proof of Theorem 5.1; if the norm diverges, the interpolation fails and the strong-type estimate does not follow.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

Share X Bluesky LinkedIn Reddit HN

If this is right

  • If Theorem 1.1 holds, the spherical trace inequality (1) follows, giving an optimal bound for orthonormal functions restricted to spheres.
  • The eigenvalue bounds of Rozenblum–Tashchiyan (number of negative eigenvalues) and Rozenblum (sum of negative eigenvalues) are recovered for Schrödinger operators with shell potentials.
  • Corollary 1.2 gives a finite bound for the interaction energy of two Frostman measures with different dimensions, complementing the Cauchy–Schwarz bound that only works when the dimensions are equal.
  • Theorem 6.1 provides Morrey-space analogues of the Lieb–Thirring and CLR inequalities, extending them to potentials in Morrey spaces.
  • The proof avoids Schatten class techniques and variational arguments, which may make the method more accessible and adaptable.

Where Pith is reading between the lines

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

  • The Fourier-analytic proof may extend to other weights or to orthonormal systems in other settings, such as metric measure spaces, though the paper sticks to Euclidean space.
  • The range p < q is likely necessary (as suggested by the discussion after Theorem 4.1), so extending the inequality to p = q would require a different mechanism.
  • The interaction energy estimate might be provable in the α = β case with a logarithmic correction, which would be a natural next step.
  • Since the proof is direct, it may produce explicit constants, which could be useful in numerical studies of shell potentials.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

0 major / 5 minor

Summary. The paper proves a fractal analogue of Lieb's Hardy–Littlewood–Sobolev inequality for orthonormal systems in L²(R^d), with the L^{q/2} norm taken against an α-Frostman measure. The main result, Theorem 1.1, gives a bound for ∑_j γ_j |D^{-s}f_j|² in L^{q/2}(dμ) in terms of the ℓ^{p/2} norm of γ and a power of the Frostman constant [μ]_α. The proof combines a Fourier-analytic reproof of Adams' fractal HLS inequality, a weak-type estimate via Bessel's inequality and a dyadic frequency split (Theorem 3.1), Lorentz refinements, and a complex interpolation step in Section 5. The paper then derives a trace inequality for orthonormal functions, an interaction energy estimate for two Frostman measures, and eigenvalue bounds for operators of the form (−Δ)^s − μ, recovering results of Rozenblum and Rozenblum–Tashchiyan. The exposition is generally clear, and the main load-bearing arguments are explicit.

Significance. If correct, Theorem 1.1 is a significant extension of Lieb's orthonormal HLS inequality to fractal measures, with explicit dependence on the Frostman constant. The approach is direct and avoids Schatten classes and variational arguments, which is a genuine methodological strength. The paper also provides new trace and eigenvalue estimates for shell potentials, and its Fourier-analytic reproof of Adams' inequality is clean. The estimates are precise and falsifiable, and the derivations are largely self-contained; the dependence on [μ]_α is tracked explicitly and no fitted constants appear. The recovery of Rozenblum-type bounds as corollaries rather than inputs is a particularly nice feature.

minor comments (5)
  1. [Section 4, proof of Theorem 4.3] In the interpolation step after Eq. (7), the statements about q0 and q1 appear reversed: choosing p0 < p gives q0 > q (not smaller) via the relation λ = α/q0 + β/p0′, and choosing p1 > p gives q1 < q (not larger). Please correct the directions or rephrase the sentence.
  2. [Section 5, interpolation display] The interpolation display '1/er = θ/2 = θ/eq0 + (1−θ)/eq1 = 1/eq' is compressed. Please spell out the definition of the source and target spaces, the interpolation parameter θ, and the choice of ep0 = 2ep/eq. In particular, verify explicitly that ep0 > eq0 is compatible with the relation 2/eq0 + (eq−2)/eq1 = 1 when eq1 is chosen large. Expanding this step would make the proof of the upgrade from weak-type to strong-type easier to check.
  3. [Section 5, Lemma 5.2] The derivation of Lemma 5.2 from Theorem 3.1 with s2 = 0 is terse. The sentence invoking the embedding ℓ^{p/2} ↪ ℓ^{q/2,1} should be expanded, for example by writing out the level-set argument; alternatively, the lemma can be derived directly from Theorem 3.2 via the embedding ℓ² ↪ ℓ^{ep}. As written, the logical link is not immediately clear.
  4. [Section 5, uniform boundedness of T_z] In the paragraph on uniform boundedness, the claim that Theorem 4.3 applies to complex (D^z)^* is justified only by the statement that the convolution kernel is 'at its largest when the kernel and function are both real and nonnegative.' Please add the explicit modulus identity | |x−y|^{-(d+z)} | = |x−y|^{-d−Re z} and use it to reduce the complex kernel to the real one.
  5. [Throughout] Minor typographical issues: 'the the number' appears in the introduction; 'inequities' should be 'inequalities'; the running title in the full text reads 'OR THONORMAL'; and in the Section 5 interpolation display, the term '1−θ/eq1' should be '(1−θ)/eq1' with a parenthesis.

Circularity Check

0 steps flagged

No circularity: main theorem proved from direct estimates; eigenvalue corollaries derived, not assumed; self-citation non-load-bearing.

full rationale

The derivation chain is self-contained. Theorem 1.1 is obtained by combining the weak-type estimate Theorem 3.1 (proved from Lemma 3.1, which is derived from first principles using Fourier decay and Frostman conditions) with the analytic interpolation in Theorem 5.1. The Lorentz-refined inequalities of Section 4 are also proved directly from Lemma 3.1 and interpolation, without invoking the target estimates. The eigenvalue bounds in Section 6 (Theorems B, C, 6.1, Corollaries 6.1–6.2) are derived from Theorems 3.1 and 3.2, not assumed; the remark that the spherical trace estimate can conversely be deduced from Rozenblum's bound is explicitly not used in the proof. The only self-citation is reference [4] as an example for Bessel's-inequality techniques, and it is not load-bearing. No fitted constants, no post-hoc exclusions, and no prediction that is equivalent by construction to an input. The compressed interpolation computation in Section 5 is terse but the exponents are determined by the stated relations, so at most a correctness-risk issue, not circularity.

Axiom & Free-Parameter Ledger

0 free parameters · 6 axioms · 0 invented entities

The proofs assume standard Fourier-analytic and interpolation facts plus the Frostman measure condition. No numerical constants are fitted to data, and no new objects are postulated.

axioms (6)
  • domain assumption μ(B(x,r)) ≤ [μ]_α r^α and ν(B(x,r)) ≤ [ν]_β r^β for all balls
    Defines the Frostman class of measures; used throughout, starting in Section 2 and in Lemma 3.1.
  • standard math The Fourier transform of |·|^{-λ} is a constant multiple of |·|^{λ-d}
    Invoked in Theorem 4.2 and in the analytic-continuation argument in Section 5.
  • standard math |(|ξ|^{-s}φ(|ξ|))^∨(x)| ≲ (1+|x|)^{-(d+1)} for smooth φ supported away from zero
    Starting point for Lemma 3.1; all the subsequent scaling estimates are built on this kernel decay.
  • standard math Marcinkiewicz real interpolation and Stein's complex interpolation apply to Lorentz spaces and vector-valued ℓ^p(L²) spaces
    Used to upgrade weak-type to strong-type estimates in Theorems 4.3 and 5.1; cited to Bergh–Löfström and Stein–Weiss.
  • standard math Bessel's inequality for orthonormal Fourier transforms: Σ_j |⟨h, f̂_j⟩|² ≤ ‖h‖₂²
    Used in the low-frequency bound in Theorem 3.1.
  • domain assumption Negative eigenvalues of (-Δ)^s+V are characterized by the Rayleigh quotient, and the span of their eigenfunctions has dimension N
    Used in Section 6 to convert the potential-energy bounds in Corollaries 6.1 and 6.2 into eigenvalue-counting bounds.

reviewed 2026-08-01 · how reviews work

0 comments
Cite this review

Pith. "Pith review of Orthonormal Sobolev estimates with fractal measures." pith.science (2026). https://pith.science/paper/R6GN6FSU

@misc{pith2026260715826,
  author       = {Pith},
  title        = {Pith review of: Orthonormal Sobolev estimates with fractal measures},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/R6GN6FSU}},
  note         = {Machine review of arXiv:2607.15826}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

We prove a fractal version of Lieb's Hardy-Littlewood-Sobolev inequality for orthonormal functions. On the one hand, this can be viewed as a trace theorem for orthonormal functions. On the other, it allows us to recover the Rozenblum-Tashchiyan bound for the number of negative eigenvalues of $-\Delta-\mu$, where $\mu$ is a shell potential. We also recover Rozenblum's bound for the sum of negative eigenvalues via a Lieb-Thirring kinetic inequality. Our proof is direct, avoiding both Schatten classes and variational arguments. We first reprove Adams' fractal Hardy-Littlewood-Sobolev inequality (for single functions) via Fourier analysis. This yields the required endpoint estimate as well as a bound for the interaction energy of Frostman measures.

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

31 extracted references · 2 linked inside Pith

  1. [1]

    D. R. Adams,A trace inequality for generalized potentials, Studia Math.48(1) (1973), 99– 105

    1973

  2. [2]

    R. D. Benguria and B. A. Loewe, Stability of Matter, inComplex Quantum Systems: Analysis of Large Coulomb Systems, IMS Lecture Notes Series, Vol. 24, World Scientific, 2013, 1–60

    2013

  3. [3]

    Bergh and J

    J. Bergh and J. L¨ ofstr¨ om,Interpolation Spaces: An Introduction, Springer-Verlag, New York, 1976

    1976

  4. [4]

    N. Bez, Y. Hong, S. Lee, S. Nakamura, and Y. Sawano,On the Strichartz estimates for orthonormal systems of initial data with regularity, Adv. Math.354(2019), 106736, 37 pp

    2019

  5. [5]

    Bilyk, D

    D. Bilyk, D. Ferizovi´ c, A. Glazyrin, R. Matzke, J. Park, and O. Vlasiuk,Potential theory with multivariate kernels, Math. Z.301(1) (2022), 341–381

    2022

  6. [6]

    Bourgain,Estimations de certaines functions maximales, C

    J. Bourgain,Estimations de certaines functions maximales, C. R. Acad. Sci. Paris310(1985), 499–502

    1985

  7. [7]

    Cwikel,Weak type estimates for singular values and the number of bound states of Schr¨ odinger operators, Ann

    M. Cwikel,Weak type estimates for singular values and the number of bound states of Schr¨ odinger operators, Ann. of Math. (2)106(1977), no. 1, 93–100

    1977

  8. [8]

    R. L. Frank,Cwikel’s theorem and the CLR inequality, J. Spectr. Theory4(2014), 1–21

    2014

  9. [9]

    R. L. Frank,Eigenvalues of Schr¨ odinger operators with complex surface potentials, Commun. Math. Phys.360(3) (2018), 1015–1030

    2018

  10. [10]

    R. L. Frank, Lieb–Thirring inequalities and other functional inequalities for orthonormal systems, inProceedings of the International Congress of Mathematicians 2022, Vol. 4, EMS Press, Berlin, 2023, 3323–3347

    2022

  11. [11]

    R. L. Frank and A. Laptev,Spectral inequalities for Schr¨ odinger operators with surface po- tentials, Spectral Theory of Differential Operators: M. Sh. Birman 80th Anniversary, Amer. Math. Soc. Transl. Ser. 2, vol. 225, Amer. Math. Soc., Providence, RI, 2008, 91–102

    2008

  12. [12]

    Garc ´ ıa-Cuerva and A

    J. Garc ´ ıa-Cuerva and A. E. Gatto,Boundedness properties of fractional integral operators associated to non-doubling measures, Studia Math.162(3) (2004), 245–261

    2004

  13. [13]

    Garc ´ ıa-Cuerva and J

    J. Garc ´ ıa-Cuerva and J. M. Martell,Two-weight norm inequalities for maximal operators and fractional integrals on non-homogeneous spaces, Indiana Univ. Math. J.50(3) (2001), 1241–1280

    2001

  14. [14]

    Grafakos,Classical Fourier Analysis, 3rd ed., Graduate Texts in Mathematics, 249, Springer, New York, 2014

    L. Grafakos,Classical Fourier Analysis, 3rd ed., Graduate Texts in Mathematics, 249, Springer, New York, 2014

    2014

  15. [15]

    Hern´ andez and J

    E. Hern´ andez and J. Soria,Spaces of Lorentz type and complex interpolation, Ark. Mat.29 (1991), 203–220

    1991

  16. [16]

    Kerman and E

    R. Kerman and E. T. Sawyer,The trace inequality and eigenvalue estimates for Schr¨ odinger operators, Ann. Inst. Fourier (Grenoble)36(1986), 207–228

    1986

  17. [17]

    M. V. Korobkov and J. Kristensen,The trace theorem, the Luzin N- and Morse–Sard proper- ties for the sharp case of Sobolev–Lorentz mappings, J. Geom. Anal.28(2018), 2834–2856

    2018

  18. [18]

    N. S. Landkof,Foundations of Modern Potential Theory, vol. 180 ofDie Grundlehren der mathematischen Wissenschaften in Einzeldarstellungen, Springer-Verlag, Berlin, 1972

    1972

  19. [19]

    E. H. Lieb,AnL p bound for the Riesz and Bessel potentials of orthonormal functions, J. Funct. Anal.51(1983), 159–165

    1983

  20. [20]

    E. H. Lieb and R. Seiringer,The stability of matter in quantum mechanics, Cambridge Uni- versity Press, Cambridge, UK, 2010

    2010

  21. [21]

    E. H. Lieb and W. E. Thirring,Bound for the Kinetic Energy of Fermions which Proves the Stability of Matter, Phys. Rev. Lett.35(11) (1975), 687–689

    1975

  22. [22]

    Maz’ya.Sobolev Spaces with Applications to Elliptic Partial Differential Equations

    V. Maz’ya.Sobolev Spaces with Applications to Elliptic Partial Differential Equations. Springer Series in Soviet Mathematics. Springer-Verlag, Berlin, 1985 (2nd edition, 2011)

    1985

  23. [23]

    Mihula, L

    Z. Mihula, L. Pick, and D. Spector,Potential trace inequalities via a Calder´ on-type theorem, J. London Math. Soc.113(3) (2026), e70504

    2026

  24. [24]

    P. T. Nam, Direct methods to Lieb–Thirring kinetic inequalities, inDensity Functionals for Many-Particle Systems: Mathematical Theory and Physical Applications of Effective Equa- tions, IMS Lecture Notes Series, Vol. 41, World Scientific, 2023, 69–97. 14 NEAL BEZ, KEITH M. ROGERS, AND SHUNYA TOYOSHIMA

    2023

  25. [25]

    Q. A. Ngˆ o,All conditions for Stein–Weiss inequalities are necessary, arXiv:2110.14220

    Pith/arXiv arXiv

  26. [26]

    P´ erez, A

    C. P´ erez, A. Seeger, B. Street, and P.-L. Yung,Two weight estimates for difference quotients, arXiv:2606.19288

    Pith/arXiv arXiv

  27. [27]

    Rozenblum,Distribution of the discrete spectrum of singular differential operators, Izv

    G. Rozenblum,Distribution of the discrete spectrum of singular differential operators, Izv. Vyssh. Uchebn. Zaved. Matematika (1976), no. 1, 75–86

    1976

  28. [28]

    Rozenblum,Lieb–Thirring estimates for singular measures, Ann

    G. Rozenblum,Lieb–Thirring estimates for singular measures, Ann. Henri Poincar´ e23(11) (2022), 4115–4130

    2022

  29. [29]

    Rozenblum and G

    G. Rozenblum and G. Tashchiyan,Eigenvalues of the Birman-Schwinger operator for singular measures: The noncritical case, J. Funct. Anal.283(11) (2022), Paper No. 109679

    2022

  30. [30]

    E. M. Stein and G. Weiss,Fractional integrals onn-dimensional Euclidean space, J. Math. Mech.7(4) (1958), 503–514

    1958

  31. [31]

    E. M. Stein and G. Weiss,Introduction to Fourier Analysis on Euclidean Spaces, Princeton Mathematical Series, Vol. 32, Princeton University Press, Princeton, NJ, 1971. Graduate School of Mathematical Sciences, The University of Tokyo, 3-8-1 Komaba, Meguro-ku, 153-8914 Tokyo, Japan Email address:bez@ms.u-tokyo.ac.jp Instituto de Ciencias Matem ´aticas CSIC...

    1971

This paper was first reviewed by deepseek-v4-flash on August 1, 2026.