Pith. sign in

REVIEW 3 major objections 4 minor 28 references

Embeddings of anisotropic Sobolev spaces into spaces of anisotropic H\"{o}lder-continuous functions

T0 review · 3 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read The paper proves that on rectangular domains, the anisotropic variable-exponent Sobolev space $W^{1,\vec p(x)}$ embeds continuously into an anisotropic Hölder space $C^{0,\vec\beta(x)}$, with each directional exponent $\beta_i$ given by…

desk verdict Lemma 2.3 is invalid in three independent places, so the main embedding theorem is unproved; the anisotropic variable-exponent Hölder space idea is worth keeping, but this paper needs a full rewrite before it merits refereeing. read the letter →

arxiv 2411.08829 v2 pith:BLDA35XE submitted 2024-11-13 math.FA math.AP

Nabil Chems Eddine , Dušan D. Repovš This is my paper · ORCID
classification math.FAmath.AP MSC 46E3546E15
keywords anisotropicvariableexponentSobolevspacesembeddingscontinuousembeddingHölder-continuousfunctionscriticalMorrey-typeestimates
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 aims to show that functions with one distributional derivative in each direction, measured in different variable Lebesgue spaces, possess direction-dependent fractional regularity on rectangular domains. Concretely, it proves a continuous embedding from the anisotropic Sobolev space $W^{1,\vec p(x)}$ into the anisotropic Hölder space $C^{0,\vec\beta(x)}$, where the Hölder exponent in direction $i$ is computed pointwise from the Sobolev exponents by $\beta_i(x) = (1 - \sum_j 1/p_j(x)) / (1 - \sum_j 1/p_j(x) + N/p_i(x))$. This would extend the classical isotropic Morrey embedding and Fan's variable-exponent results to a setting where smoothness can differ along different axes, which matters for models of composite materials, turbulent flows, and image restoration. The proof works through a local oscillation estimate on cubes and a covering argument.

What carries the argument

The load-bearing object is the anisotropic variable-exponent Hölder space $C^{0,\vec\beta(x)}$ with norm (2.1): a function is measured by its sup norm plus the supremum over pairs $x,y$ of $|u(x)-u(y)| / \sum_i |x_i-y_i|^{\beta_i(x,y)}$. The proof of the embedding rests on Lemma 2.3, a local Morrey-type estimate on a unit cube: $|u(x)-u(y)| \le c\|u\| \sum_i |x_i-y_i|^{\beta_i(x,y)}$. To prove it, the authors introduce a small translated box $\Omega(s)$ with side lengths $s^{1/\beta_i^-}$, where $s = \sum |x_i-y_i|^{\beta_i(x,y)}$, use the fundamental theorem of calculus along segments, Fubini, and Hölder's inequality to control the average oscillation, and then patch cubes across the rectangular domain.

What would settle it

Take $N=2$, $Q=(0,1)^2$, exponents $p_1(x)=4+0.5\sin(2\pi x_1)$ and $p_2(x)=4+0.5\cos(2\pi x_2)$, so $p_m>N$ and $\beta_1$ varies with $x$. Choose $x=(0.25,0.5)$ and $y=(0.35,0.5)$, so only the first coordinate differs. Compute $s=|x_1-y_1|^{\beta_1(x,y)}$ and compare $s^{1/\beta_1^-}$ with $|x_1-y_1|$: since $\beta_1(x,y)>\beta_1^-$, the box side is smaller than the coordinate gap, so $y$ lies outside the asserted box $\Omega(s)$. Directly evaluating the Lemma 2.3 inequality for a smooth function along this pair determines whether the claimed constant $c$ can exist and therefore whether the proof's main estimate holds.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1.1: for a rectangular domain $\Omega \subset \mathbb{R}^N$ and continuous exponents $p_i$ with $N < p_m(x) \le p_M(x) < \infty$, the space $W^{1,\vec p(x)}(\Omega)$ is continuously embedded in $C^{0,\vec\beta(x)}(\Omega)$, where $\beta_i(x)$ is given by the formula above. The embedding is witnessed by the pointwise bound $|u(x)-u(y)| \le C\|u\|_{W^{1,\vec p(x)}} \sum_i |x_i-y_i|^{\beta_i(x,y)}$ with $\beta_i(x,y)=\min(\beta_i(x),\beta_i(y))$. Two corollaries are drawn: the isotropic case $p_i=p$ reduces $\beta$ to $1-N/p(x)$, and larger exponents $p_i$ yield larger Hölder exponents, approaching Lipschitz regularity in the limit.

Load-bearing premise

The proof's key local estimate rests on the assumption that a small anisotropic distance forces the two points into a box whose side lengths are set by the smallest Hölder exponent, and that Hölder's inequality can be applied with the exponent frozen at one point; if either premise fails for genuinely variable exponents, the estimate and hence the embedding proof collapse.

Editorial extensions

If this is right

  • Every element of $W^{1,\vec p(x)}(\Omega)$ on a rectangle has a continuous representative whose oscillation along coordinate direction $i$ is controlled by the local exponent $\beta_i(x,y)$, giving quantitative anisotropic regularity rather than a single global modulus.
  • When the Sobolev exponents are all equal to $p(x)$, the embedding reduces to the known isotropic one with Hölder exponent $1 - N/p(x)$, so the theorem contains the classical scalar case as a special case.
  • Raising the Sobolev exponent in one direction raises the Hölder exponent in that direction, so the result interpolates between mere continuity and Lipschitz continuity as $p_i$ increases.
  • The embedding supplies a Banach-space framework for studying anisotropic variable-exponent PDEs on rectangular domains: solutions lying in $W^{1,\vec p(x)}$ automatically have the directional Hölder regularity needed in compactness arguments.

Reading between the lines

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

  • The pointwise estimate implies a scale-invariant bound: after rescaling a rectangular domain, the embedding constant depends only on the anisotropy of the exponents and not on the domain diameter, so the same argument applies to cubes of arbitrary side length with adjusted constants.
  • One could test the sharpness of the exponent formula numerically: approximate functions such as $|x_i|^{\alpha}$ on a rectangle and compare the largest achievable directional Hölder exponent with the predicted $\beta_i$; agreement would support the formula, while disagreement would suggest a different critical exponent.
  • The counterexamples in Section 4 suggest that rectangularity is not merely a technical convenience: if a similar embedding holds on cuspidal domains, it would need a modified norm or boundary-dependent exponents, and characterizing that is a natural next step.
  • The anisotropic Hölder norm could serve as a regularity diagnostic in applied fields: the ratio $\beta_i/\beta_j$ measures how much smoother a signal or velocity field is along one axis than another, which is exactly the directional information image-restoration and turbulence models need.
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

3 major / 4 minor

Summary. The paper defines an anisotropic variable-exponent H\"older space C^{0,\vec{\beta}(x)} on rectangular domains and claims a continuous embedding W^{1,\vec{p}(x)} \hookrightarrow C^{0,\vec{\beta}(x)} with explicit exponents \beta_i given by (2.2), under the assumption that each p_i is continuous, bounded away from 1 and infinity, and satisfies p_i(x) > N. The proof combines Fan's embedding theorem with a Morrey-type estimate (Lemma 2.3) intended to control |u(x)-u(y)| by \sum_i |x_i-y_i|^{\beta_i(x,y)}. Corollaries address the isotropic reduction and monotonicity of the exponents; Section 4 gives informal examples on cusped domains, and Section 5 presents qualitative applications to heat conduction and porous-media flow.

Significance. If valid, the main theorem would be a useful generalization of Fan's isotropic variable-exponent embedding and of R\'akosn\'ik's constant-exponent anisotropic results, providing explicit, direction-dependent H\"older exponents. The definition of the anisotropic H\"older space through the pointwise minimum \beta_i(x,y)=\min\{\beta_i(x),\beta_i(y)\} is natural, and the paper contains no fitted parameters or circular normalizations; the claimed embedding is a genuine, falsifiable statement. However, the central estimate in Lemma 2.3 is false as written: the decisive exponent comparison fails on a concrete two-point example, and the proof also applies H\"older's inequality as if the exponent were constant. Because Lemma 2.3 is the only mechanism producing the pointwise H\"older bound, the main theorem is not established by the arguments given.

major comments (3)
  1. [Lemma 2.3, Eq. (2.3) and following inequalities] The step replacing s^{1/\beta_i^- - \sigma/p_i(x)} by s, i.e. asserting 1/\beta_i^- - \sigma/p_i(x) \ge 1, is false in general. Take N=2, p_1(x)=3, p_2(x)=100, p_1(y)=100, p_2(y)=3. Using (2.2) with the denominator interpreted as N/p_i(x) (as Theorem 1.1 requires), one gets \beta_1(x)=\beta_2(y)\approx 0.496 and \beta_1(y)=\beta_2(x)\approx 0.970, so \beta_1^-=\beta_2^-\approx 0.496 and \sigma=\sum_i 1/\beta_i^-\approx 4.03. For i=1 at x, the exponent is 1/0.496 - 4.03/3 \approx 0.67 < 1, so the estimated factor is s^{0.67}, which is not dominated by s as s\to 0. Since Lemma 2.3 is the only mechanism giving the pointwise H\"older estimate, Theorem 1.1 is not proved. The translated-box containment is not the difficulty: \beta_i^-\le \beta_i(x,y) gives s^{1/\beta_i^-}\ge |x_i-y_i|, so a translated box of side s^{1/\beta_i^-} can indeed contain both points.
  2. [Lemma 2.3, after Eq. (2.3)] H\"older's inequality is applied as \int_{\Omega(s,t)} |\partial_i u|\,dz \le \|\partial_i u\|_{L^{p_i(x)}(Q)}\,(\operatorname{meas}\Omega(s,t))^{1-1/p_i(x)}, which treats p_i(x) as a constant exponent. In variable-exponent Lebesgue spaces the dual norm of the characteristic function is not generally (meas)^{1-1/p_i(x)}; the correct estimate involves the norm in L^{p_i'(\cdot)} and depends on the log-H\"older constant or on global bounds in a more subtle way. This step is therefore unjustified even if the exponent comparison in the previous comment were repaired.
  3. [Lemma 2.1 and Theorem 1.1] Lemma 2.1 is stated for \vec p\in(C_+(\Omega))^N only, but the introduction describes Fan's Theorem 2.5 as requiring the log-H\"older condition \vec p\in(C_{\log}^+(\Omega))^N. The proof of Theorem 1.1 invokes Lemma 2.1 at (3.1) and again in Case 1 of Lemma 2.3 to obtain continuity and the s\ge 1 estimate. Under the stated hypotheses, that embedding into C(\Omega) is not justified by the cited theorem. The authors should either add the log-H\"older condition to the hypotheses or provide a proof of Lemma 2.1 as stated.
minor comments (4)
  1. [Equation (2.2)] The denominator in (2.2) reads 1-\sum_j 1/p_j(x)+N/p_1(x) for every i, which would make all \beta_i identical and contradict the anisotropic nature of the claimed embedding; it should presumably be N/p_i(x).
  2. [Section 4] The examples in Section 4 are informal and do not constitute rigorous counterexamples to any stated theorem: for instance, the function u(x,y)=\sqrt{x} on the cusp domain is not shown to belong to W^{1,4}(\Omega), nor is it shown to fail every H\"older condition.
  3. [Corollary 3.2] Corollary 3.2 is not a precise mathematical statement: saying that \beta_i(x) approaches 1 as p_i(x)\to\infty does not formulate a comparative embedding statement, and the proof does not establish monotonicity as a theorem about the embedding.
  4. [References and text] There are numerous typographical and notational inconsistencies (e.g., C^+ vs. C_{\log}^+, L^{p_i(x)}(Q) vs. L^{p_i(\cdot)}(Q), and missing spaces in displayed formulas); a careful revision is needed if the paper is resubmitted.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the main embedding is derived from Fan's external embedding theorem and elementary estimates, with no fitted inputs or load-bearing self-citations.

full rationale

No specific circular step is present. Theorem 1.1 is proved by combining Lemma 2.1, which is an external result of Fan [9], with Lemma 2.3, whose proof uses the fundamental theorem of calculus, Fubini's theorem, and Hölder's inequality. The target Hölder exponent vector β is explicitly prescribed by equation (2.2) rather than fitted to the conclusion, and the space C^{0,→β(x)} in Definition 2.2 is defined independently of the embedding being proved. The authors' own cited works appear only as background or applications, not as the source of any load-bearing assertion. Section 4 explicitly identifies limitations of the theorem for non-rectangular domains, which further indicates that the claim is not being forced by a self-supporting construction. Possible mathematical gaps in the exponent estimates of Lemma 2.3 are correctness concerns, not circularity, and the honest finding is therefore no significant circularity.

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

The central proof rests on Fan's embedding, the variable-exponent Hölder inequality, density of smooth functions, and the questionable box-containment assertion. The theorem statement omits the log-Hölder condition needed by Fan's result.

assumptions (4)
  • domain assumption Fan's Lemma 2.1: W^{1,→p(x)}(Ω) embeds continuously into C(Ω) for rectangular Ω when p_i are log-Hölder continuous and N < p_m(x).
    Invoked in Lemma 2.3 and Theorem 1.1, but Theorem 1.1 only assumes p_i ∈ C_+(Ω), not log-Hölder. The proof depends on Fan's theorem, which requires log-Hölder.
  • domain assumption Variable-exponent Hölder inequality: ∫ |fg| ≤ C ||f||_{p(⋅)} ||g||_{p'(⋅)} with C depending on the log-Hölder constant.
    Used in Lemma 2.3 to bound integrals over Ω(s,t), but applied as if p_i(x) were a constant exponent, which is not valid.
  • ad hoc to paper The translated box Ω(s) with side lengths s^{1/β_i^-} contains both x and y.
    Asserted in Lemma 2.3 but false in general when β_i(x,y) > β_i^-, because |x_i-y_i| can exceed s^{1/β_i^-}.
  • domain assumption Density of C∞ in W^{1,→p(x)}(Ω) for rectangular domains.
    The proof of Theorem 1.1 argues for u ∈ C∞ and concludes by density, but density is not stated or cited in this context.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Embeddings of anisotropic Sobolev spaces into spaces of anisotropic H\"{o}lder-continuous functions." pith.science (2026). https://pith.science/paper/BLDA35XE

@misc{pith2026241108829,
  author       = {Pith},
  title        = {Pith review of: Embeddings of anisotropic Sobolev spaces into spaces of anisotropic H\"older-continuous functions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BLDA35XE}},
  note         = {Machine review of arXiv:2411.08829}
}
read the original abstract

We introduce a novel framework for embedding anisotropic variable exponent Sobolev spaces into spaces of anisotropic variable exponent H\"{o}lder-continuous functions within rectangular domains. We establish a foundational approach to extend the concept of H\"{o}lder continuity to anisotropic settings with variable exponents, providing deeper insight into the regularity of functions across different directions. Our results not only broaden the understanding of anisotropic function spaces but also open new avenues for applications in mathematical and applied sciences.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

28 extracted references · 27 canonical work pages

  1. [1]

    Acerbi, N

    E. Acerbi, N. Fusco; Partial regularity under anisotropic(p, q) growth conditions, J. Differ. Equ. 107 (1994), no. 1, 46-67, DOI: https://doi.org/10.1006/jdeq.1994.1002

    arXiv 1994

  2. [2]

    R. A. Adams; Sobolev Spaces, Academic Press, New Y ork, 1975

    work page 1975

  3. [3]

    Chems Eddine, M

    N. Chems Eddine, M. A. Ragusa, D. D. Repovˇ s; On the concentration-compactness principle for anisotrop ic variable exponent Sobolev spaces and its applications , Fract. Calc. Appl. Anal. 27 (2024), 725–756, DOI: https://doi.org/10.1007/s13540-024-00246-8

    work page doi:10.1007/s13540-024-00246-8 2024

  4. [4]

    Chems Eddine, D

    N. Chems Eddine, D. D. Repovˇ s; The Neumann problem for a class of generalized Kirchhoff-type potential systems , Bound. V alue. Probl. 19 (2023), no. 1, art. 19, DOI: https://doi.org/10.1186/s13661-023-01705-6

    work page doi:10.1186/s13661-023-01705-6 2023

  5. [5]

    Ciani, V

    S. Ciani, V . V espri; On H ¨older continuity and equivalent formulation of intrinsic H arnack estimates for an anisotropic parabolic degenerate prototype equation , Constr. Math. Anal. 4 (2021), no. 1, 93-103, DOI: https://doi.org/10.33205/cma.824336

    work page doi:10.33205/cma.824336 2021

  6. [6]

    Cruz-Uribe and A

    D. Cruz-Uribe and A. Fiorenza; V ariable Lebesgue Spaces: F oundations and Harmonic Analysis, Birkh¨ auser, Basel, 2013

    work page 2013

  7. [7]

    Diening, P

    L. Diening, P . Harjulehto, P . H¨ ast¨ o, M. R˚ uˇ ziˇ cka;Lebesgue and Sobolev Spaces with V ariable Exponents , Lecture Notes in Mathematics, Springer-V erlag, Heidelberg, 2011

    work page 2011

  8. [8]

    D. E. Edmunds, J. Rakosnik; Sobolev embeddings with variable exponent , Stud. Math. 143 (2000), no. 3, 267-293

    work page 2000

Show all 28 references
  1. [9]

    Fan; Anisotropic variable exponent Sobolev spaces and ⃗p(·)-Laplacian equations , Complex V ar

    X. Fan; Anisotropic variable exponent Sobolev spaces and ⃗p(·)-Laplacian equations , Complex V ar. Elliptic Equ. 56 (2010), no. 7-9, 623-642, DOI: https://doi.org/10.1080/17476931003728412

    2010 doi

  2. [10]

    X. Fan, D. Zhao; On the spaces L p(x)(Ω ) and W m,p(x)(Ω ), J. Math. Anal. Appl. 263 (2001), no. 2, 424-446, DOI: https://doi.org/10.1006/jmaa.2000.7617

    2001

  3. [11]

    Gutlyanskii, O

    V . Gutlyanskii, O. Nesmelova, V . Ryazanov , E. Y akubov; T oward the theory of semi-linear Beltrami equa- tions, Constr. Math. Anal. 6 (2023), no. 3, 151-163, DOI: https://doi.org/10.33205/cma.1248692. 12 N. CHEMS EDDINE, D.D. REPOV ˇS

    2023 doi

  4. [12]

    Haˇ skovec, C

    J. Haˇ skovec, C. Schmeiser; A note on the anisotropic generalizations of the Sobolev and Morrey embedding theorems, Monatsh. Math. 158 (2009), 71-79, DOI: https://doi.org/10.1007/s00605-008-0059-x

    2009 doi

  5. [13]

    C. B. Morrey, Jr.; Functions of several variables and absolute continuity, II , Duke J. Math. 6 (1940), no. 1, 187-215, DOI: https://doi.org/10.1215/S0012-7094-40-00615-9

    1940 doi

  6. [14]

    C. B. Morrey, Jr.; Multiple integral problems in the calculus of variations an d related topics, Ann. Sc. Norm. Super. Pisa Cl. Sci. 14 (1960) no. 1, 1–61. http://www.numdam.org/item/ASNSP_1960_3_14_1_1_0

    1960

  7. [15]

    N. S. Papageorgiou, V . D. R˘ adulescu, D. D. Repovˇ s;Nonlinear Analysis - Theory and Applications , Springer, Cham, 2019

    2019

  8. [16]

    V . D. R˘ adulescu, D. D. Repovˇ s;Partial Differential Equations with V ariable Exponents: Variational Methods and Qualitative Analysis, CRC Press, Boca Raton, FL, 2015

    2015

  9. [17]

    R´ akosn´ ık;Some remarks to anisotropic Sobolev Spaces I , Beitr¨ age Anal

    J. R´ akosn´ ık;Some remarks to anisotropic Sobolev Spaces I , Beitr¨ age Anal. (1979), no. 13, 55-68

    1979

  10. [18]

    R´ akosn´ ık;Some remarks to anisotropic Sobolev spaces II , Beitr¨ age Anal

    J. R´ akosn´ ık;Some remarks to anisotropic Sobolev spaces II , Beitr¨ age Anal. (1981), no. 15, 127-140

    1981

  11. [19]

    S. L. Sobolev; On a theorem of functional analysis , Mat. Sb. [Math USSR-Sb.], 4 (1938), no. 3, 471-497

    1938

  12. [20]

    Troisi; T eoremi di inclusione per spazi di Sobolev non isotropi, Ricerche Mat

    M. Troisi; T eoremi di inclusione per spazi di Sobolev non isotropi, Ricerche Mat. 18 (1969), no. 3, 3-24

    1969

  13. [21]

    Troisi; Ulteriori contributi alla teoria degli spazi di Sobolev non isotropi, Ricerche Mat

    M. Troisi; Ulteriori contributi alla teoria degli spazi di Sobolev non isotropi, Ricerche Mat. 20 (1971), no. 1, 90-117

    1971

  14. [22]

    K. K. Kar; Composite Materials: Processing, Applications, Characte rizations, 1st ed., Springer, Berlin Hei- delberg, New Y ork, NY , 2016

    2016

  15. [23]

    P . A. Davidson; Turbulence: An Introduction for Scientists and Engineers , Oxford University Press, Oxford, 2015

    2015

  16. [24]

    T. K. Moon, W . Stirling; Mathematical Methods And Algorithms F or Signal Processing, Prentice Hall, Upper Saddle River, NJ, 1999

    1999

  17. [25]

    R. C. Gonzalez, R. E. Woods; Digital Image Processing , 3rd ed., Prentice-Hall, Upper Saddle River, NJ, 2007

    2007

  18. [26]

    Z. Ren, F. Fang, N. Y an et al.;State of the art in defect detection based on machine vision, Int. J. of Precis. Eng. and Manuf.-Green Tech. 9 (2022), 661-691. DOI: https://doi.org/10.1007/s40684-021-00343-6

    2022 doi

  19. [27]

    Johansen-Berg, T

    H. Johansen-Berg, T. E. J. Behrens; Diffusion MRI: From Quantitative Measurement to In Vivo Neu - roanatomy, Elsevier, London, 2013

    2013

  20. [28]

    K. K. Shung, M. Smith, B. M. Tsui; Principles of Medical Imaging , Academic Press, New Y ork, 2012

    2012

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.