pith. sign in

arxiv: 2605.22265 · v3 · pith:WP3DTEIPnew · submitted 2026-05-21 · 🧮 math.DG · math.AT· math.PR· math.ST· stat.TH

Empirical Hodge Laplacians: Spectral Convergence and Harmonic Forms from Point Clouds

H\^ong V\^an L\^e This is my paper

Pith reviewed 2026-06-30 16:27 UTC · model grok-4.3

classification 🧮 math.DG math.ATmath.PRmath.STstat.TH
keywords Hodge Laplacianpoint cloudsspectral convergenceBetti numbersharmonic formsMosco convergencemanifold reconstructioncharacteristic classes
0
0 comments X

The pith

Point cloud samples on a smooth manifold yield empirical Hodge operators whose low-lying spectrum recovers the Betti numbers and converges to the harmonic forms.

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

The paper constructs deformed Hodge Laplacians from the extrinsic geometry of a closed Riemannian submanifold and shows these converge uniformly to the classical Hodge Laplacian as the deformation parameter tends to zero. For an i.i.d. uniform point cloud of size m, the corresponding empirical operators, when the parameter is scaled as m to the power minus one over twice the dimension, converge uniformly in probability and satisfy compact Mosco convergence of their quadratic forms. This convergence forces the cluster of eigenvalues near zero to have multiplicity exactly equal to the k-th Betti number and the associated eigenvectors to converge to the space of harmonic k-forms. The same construction supplies consistent estimators for the second fundamental form, curvature tensor, and Weitzenbock endomorphisms, so that characteristic numbers become recoverable from samples alone.

Core claim

Under the scaling t = m^{-1/(2n)}, the empirical Hodge operators satisfy uniform consistency in probability and compact Mosco convergence; consequently the empirical spectral cluster near zero contains exactly b_k eigenvalues, counted with multiplicity, and converges in the transported discrete L^2 sense to the space of harmonic k-forms. Consistent empirical estimators are also obtained for the tangent projection, second fundamental form, Riemannian curvature tensor, and Weitzenbock curvature endomorphisms, allowing recovery of Betti numbers, harmonic representatives, Pontryagin forms, and Pontryagin numbers from the point cloud.

What carries the argument

The family of deformed Hodge Laplacians Δ_t^k defined via extrinsic geometry of the embedding, together with their empirical point-cloud versions δ̂_{t,S_m}^k that achieve Mosco convergence under the scaling t = m^{-1/(2n)}.

If this is right

  • The multiplicity of the near-zero cluster in the empirical spectrum equals the Betti number b_k for each degree k.
  • The corresponding empirical eigenvectors converge to harmonic k-forms and therefore furnish consistent representatives of de Rham cohomology.
  • Pontryagin numbers and other characteristic classes computed from the estimated curvature endomorphisms converge to their manifold values.
  • The same point-cloud construction recovers the tangent projection and second fundamental form consistently.

Where Pith is reading between the lines

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

  • The convergence statements could be checked numerically on synthetic samples from the torus or projective plane, where the Betti numbers are known exactly.
  • The framework might extend to point clouds with additive noise or non-uniform sampling densities if the scaling and consistency proofs can be adapted.
  • Because the method produces both topological invariants and curvature estimators, it supplies a differential-geometric counterpart to existing persistent-homology pipelines on point clouds.

Load-bearing premise

The manifold is C^4-smooth, closed, connected, orientable and embedded in Euclidean space, while the points are drawn i.i.d. uniformly at random from it.

What would settle it

Sample a large point cloud from the standard 2-sphere, form the empirical operator at the prescribed scaling for k=0, and verify whether exactly one eigenvalue lies near zero while the remainder are bounded away from zero by a positive gap independent of sample size.

read the original abstract

Let $M^n\subset\mathbb R^d$ be a closed, connected, orientable $C^4$-smooth Riemannian submanifold of dimension $n\ge3$. We construct, for each degree $0\le k\le n$, a family of deformed Hodge Laplacians $\Delta_t^k$, $t>0$, defined in terms of the extrinsic geometry of $M^n$, and prove that $\Delta_t^k$ converges uniformly to the classical Hodge Laplacian $\Delta^k$ as $t\to0^+$. Given an i.i.d.\ uniformly distributed point cloud $S_m\subset M^n$, we define empirical Hodge operators $\widehat\Delta_{t,S_m}^k$. Under the scaling $t=m^{-1/(2n)}$, we prove uniform consistency in probability and compact Mosco convergence of the associated quadratic forms. Consequently, the empirical spectral cluster near zero contains exactly the $k$-th Betti number $b_k$ of eigenvalues, counted with multiplicity, and converges in the transported discrete $L^2$-sense to the space of harmonic $k$-forms. We also construct consistent empirical estimators of the tangent projection, the second fundamental form, the Riemannian curvature tensor, and the Weitzenb\"ock curvature endomorphisms. As applications, we obtain consistent recovery of the Betti numbers and harmonic representatives of de Rham cohomology, as well as of the Pontryagin forms, characteristic classes, and Pontryagin numbers of $M^n$ from sampled data.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 3 minor

Summary. The paper constructs a family of deformed Hodge Laplacians Δ_t^k (0 ≤ k ≤ n) on a closed C^4 Riemannian submanifold M^n ⊂ R^d using extrinsic geometric quantities, proves uniform convergence Δ_t^k → Δ^k as t → 0^+, and for i.i.d. uniform point clouds S_m of size m defines empirical operators widehatΔ_{t,S_m}^k. Under the explicit scaling t = m^{-1/(2n)}, it establishes uniform consistency in probability together with compact Mosco convergence of the associated quadratic forms; this implies that the empirical spectral cluster near zero consists of exactly b_k eigenvalues (with multiplicity) and that the corresponding eigenspaces converge in the transported discrete L^2 sense to the space of harmonic k-forms. Consistent estimators are also constructed for the tangent projection, second fundamental form, curvature tensor, and Weitzenböck endomorphisms, yielding consistent recovery of Betti numbers, harmonic representatives, Pontryagin forms, and characteristic numbers.

Significance. If the stated convergence results hold, the work supplies a rigorous bridge between extrinsic differential geometry and spectral methods on point clouds, with explicit scaling, Mosco convergence, and recovery of topological invariants (b_k and harmonic forms) as well as geometric quantities. The parameter-free choice of scaling and the use of standard Γ-convergence arguments to pass from pointwise kernel approximations to spectral convergence are notable strengths that could support further applications in manifold learning and topological data analysis.

minor comments (3)
  1. [Abstract] Abstract, line on uniform consistency: the precise norm in which consistency holds (e.g., operator norm on L^2 or pointwise) is not stated; this should be made explicit already in the abstract and introduction.
  2. [Setup] The transported discrete L^2 inner product used for eigenspace convergence is referenced but its precise definition and the construction of the transport maps appear only after the main theorems; moving a brief definition to the setup section would improve readability.
  3. The C^4 regularity assumption on M is used for the kernel approximations of curvature quantities; a remark on whether C^3 would suffice for the Mosco convergence step (or why C^4 is necessary) would clarify the sharpness of the hypotheses.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive summary of our work, the assessment of its significance, and the recommendation for minor revision. The report lists no major comments.

Circularity Check

0 steps flagged

No significant circularity identified

full rationale

The paper's central results establish uniform consistency and compact Mosco convergence of empirical Hodge operators to the classical Hodge Laplacian under the explicitly chosen scaling t = m^{-1/(2n)}, using kernel approximations for extrinsic geometry and standard Γ-convergence arguments. These steps derive the spectral cluster multiplicity equaling b_k and convergence of eigenspaces without reducing any prediction to a fitted input, self-definition, or load-bearing self-citation. All estimators for tangent projection, second fundamental form, curvature, and Weitzenböck terms are constructed directly from the point cloud and shown consistent via the same limit process, keeping the derivation self-contained against external manifold geometry.

Axiom & Free-Parameter Ledger

1 free parameters · 2 axioms · 0 invented entities

The central claims rest on the C^4 smoothness of the manifold, the i.i.d. uniform sampling assumption, and standard results from Riemannian geometry and spectral theory; no new entities are postulated.

free parameters (1)
  • deformation scale t
    Chosen as t = m^{-1/(2n)} to balance bias and variance in the empirical operator; this is an explicit rate rather than a fitted constant.
axioms (2)
  • domain assumption M is a closed connected orientable C^4 Riemannian submanifold of dimension n ≥ 3
    Invoked to guarantee existence of the classical Hodge Laplacian and the convergence statements.
  • domain assumption Point cloud S_m is i.i.d. uniformly distributed on M
    Required for the law of large numbers underlying uniform consistency in probability.

pith-pipeline@v0.9.1-grok · 5810 in / 1380 out tokens · 31350 ms · 2026-06-30T16:27:10.877917+00:00 · methodology

Review history (2 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

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

  1. K\"ahler thresholds

    math.AG 2026-06 unverdicted novelty 7.0

    Proves even odd Betti numbers and positive even Betti numbers for compact acK manifolds using compactness and Sobolev Hodge theory.

Reference graph

Works this paper leans on

43 extracted references · 8 canonical work pages · cited by 1 Pith paper

  1. [1]

    S. B. Alexander, R. L. Bishop, Gauss equation and injectivity radii for subspaces in spaces of curvature bounded above. Geom. Dedicata 117(2006), 65–84

    2006

  2. [2]

    Aamari, C

    E. Aamari, C. Levrard, Nonasymptotic rates for manifold, tangent space and cur- vature estimation. Ann. Statist. 47(1): 177-204 (2019). DOI: 10.1214/18-AOS1685, long versionhttps://doi.org/10.1214/18-AOS1685

    work page doi:10.1214/18-aos1685 2019

  3. [3]

    Aamari, C

    E. Aamari, C. Levrard, Supplementary Material for Nonasymptotic rates for manifold, tangent space and curvature estimation.https: //projecteuclid.org/journals/annals-of-statistics/volume-47/issue-1/ Nonasymptotic-rates-for-manifold-tangent-space-and-curvature-estimation/ 10.1214/18-AOS1685.full?tab=ArticleLinkSupplemental

    work page doi:10.1214/18-aos1685.full

  4. [4]

    Bengio, O

    Y. Bengio, O. Delalleau, N. Le Roux, J.-F. Paiement, P. Vincent, and M. Ouimet. Out-of-sample extensions for LLE, Isomap, MDS, Eigenmaps, and Spectral Clus- tering. Advances in Neural Information Processing Systems 16 (NIPS 2003), 177– 184 (2004)

    2003

  5. [5]

    Berline, E

    N. Berline, E. Getzler, M. Vergne, Heat Kernels and Dirac Operators. 2nd Edition, Springer, 1996

    1996

  6. [6]

    Belkin, P

    M. Belkin, P. Niyogi, Laplacian eigenmaps for dimensionality reduction and data representation. Neural computation, 15(6), 1373–1396 (2003)

    2003

  7. [7]

    Belkin, P

    M. Belkin, P. Niyogi, Towards a theoretical foundation for Laplacian-based mani- fold methods. Journal of Computer and System Sciences, 74(8), 1289–1308 (2008)

    2008

  8. [8]

    Belkin, P

    M. Belkin, P. Niyogi, Convergence of Laplacian Eigenmaps. Ad- vances in Neural Information Processing Systems 19 (NIPS 2006), 129-136https://proceedings.neurips.cc/paper_files/paper/2006/ file/5848ad959570f87753a60ce8be1567f3-Paper.pdf, long version: https://misha.belkin-wang.org/papers/CLEM_08.pdf

    2006

  9. [9]

    Calder, and N

    J. Calder, and N. Garc´ ıa Trillos, Improved spectral convergence rates for graph Laplacians onε-graphs andk-NN graphs. Applied and Computational Harmonic Analysis, 60(2022), 123-175

    2022

  10. [10]

    Y. Cao, D. Li, H. Sun, et al., Efficient Weingarten map and curvature estimation on manifolds. Mach Learn. 110 (2021), 1319–1344.https://doi.org/10.1007/ s10994-021-05953-4

    2021

  11. [11]

    K.T., Chen, Iterated path integrals. Bull. Am. Math. Soc. 83 (1977), 831–879

    1977

  12. [12]

    Davis, W.M

    C. Davis, W.M. Kahan, The rotation of eigenvectors by a perturbation. III, SIAM Journal on Numerical Analysis, vol. 7 (1970) N 1, 1–46

    1970

  13. [13]

    Dey, A.N

    T.K. Dey, A.N. Hirani, and B. Krishnamoorthy, Optimal homologous cycles, to- tal unimodularity, and linear programming. SIAM Journal on Computing, 40, N4(2011), SIAM, 1026–1044

    2011

  14. [14]

    Dudley, Uniform Central Limit Theorems, Cambridge University Press, Cambridge, UK, 2014

    R.M. Dudley, Uniform Central Limit Theorems, Cambridge University Press, Cambridge, UK, 2014

    2014

  15. [15]

    D. B. Dunson, H.-T. Wu, and N. Wu, Spectral convergence of graph Laplacian and heat kernel reconstruction inL ∞ from random samples. Applied and Com- putational Harmonic Analysis, 55(2021), 282-336

    2021

  16. [16]

    Federer, Geometric Measure Theory, Die Grundlehren der mathematischen Wissenschaften, 153 (1969), Springer-Verlag, New York

    H. Federer, Geometric Measure Theory, Die Grundlehren der mathematischen Wissenschaften, 153 (1969), Springer-Verlag, New York

    1969

  17. [17]

    Federer, Curvature measures

    H. Federer, Curvature measures. Trans. Amer. Math. Soc. 93(1959) 418–491

    1959

  18. [18]

    Fiorenza, K

    D. Fiorenza, K. Kawai, H. V. Lˆ e and L. Schwachh¨ ofer, Almost formality of mani- folds of low dimension. Ann. Sc. Norm. Super. Pisa Cl. Sci. (5), vol. XXII (2021), 79-107. 48 H ˆONG V ˆAN L ˆE

    2021

  19. [19]

    Fiorenza, H

    D. Fiorenza, H. V. Lˆ e, UnitalC ∞-algebras and the real homotopy type of (r−1)-connected compact manifolds of dimension≤ℓ(r−1) + 2, https://arxiv.org/abs/2310.19506, Ann. Sc. Norm. Super. Pisa,https://doi. org/10.2422/2036-2145.202401_0032025

    work page doi:10.2422/2036-2145.202401_0032025 2036

  20. [20]

    Gin´ e, A

    E. Gin´ e, A. Guillou, On consistency of kernel density estimators for randomly censored data: rates holding uniformly over adaptive intervals. Ann. Inst. Henri Poincar´ e, Probabilit´ es et Statistiques, 37(2001) 503–522

    2001

  21. [21]

    Gin´ e, A

    E. Gin´ e, A. Guillou, Rates of strong uniform consistency for multivariate kernel density estimators. Annales de l’Institut Henri Poincar´ e, Probabilit´ es et Statis- tiques, Vol. 38, No. 6, (2002), pp. 907–921

    2002

  22. [22]

    G. H. Golub, C. F. Van Loan. Matrix Computations, 4th edition, Johns Hopkins University Press, (2013)

    2013

  23. [23]

    Gray, The volume of a small geodesic ball of a Riemannian manifold

    A. Gray, The volume of a small geodesic ball of a Riemannian manifold. Michigan Mathematical Journal, vol.20, no.4, 329–344 (1974)

    1974

  24. [24]

    R. A. Horn, C. R. Johnson, Matrix analysis (2nd ed.). Cambridge University Press. (2012)

    2012

  25. [25]

    Jost, Riemannian Geometry and Geometric Analysis

    J. Jost, Riemannian Geometry and Geometric Analysis. 7th Edition, Springer, 2017

    2017

  26. [26]

    Kato, Perturbation Theory for Linear Operators

    T. Kato, Perturbation Theory for Linear Operators. (Classics in Mathematics), Springer (1995)

    1995

  27. [27]

    , A. V. Knyazev, Toward the optimal preconditioned eigensolver: Locally opti- mal block preconditioned conjugate gradient method, SIAM Journal on Scientific Computing, 23(2001), Nr. 2, 517–541,

    2001

  28. [28]

    Kobayashi, K

    S. Kobayashi, K. Nomizu, Foundations of Differential Geometry. vol. I, Inter- science Publishers, 1963

    1963

  29. [29]

    Kobayashi, K

    S. Kobayashi, K. Nomizu, Foundations of Differential Geometry. vol. II, Inter- science Publishers, 1969

    1969

  30. [30]

    C. Lanczos, An iteration method for the solution of the eigenvalue problem of lin- ear differential and integral operators, Journal of Research of the National Bureau of Standards, 45 (1950), 4, 255–282,

    1950

  31. [31]

    H. V. Lˆ e, Minimal Unital CyclicC∞-Algebras and the Real and Rational Homo- topy Type of Closed Manifolds.https://arxiv.org/abs/2603.01219

    work page arXiv

  32. [32]

    H. V. Lˆ e, H. Q. Minh, F. Protin, W. Tuschmann, Mathematical Foundations of Machine Learning, Springer 2026 (to appear)

    2026

  33. [33]

    S. A. Merkulov, Strong homotopy algebras of a K¨ ahler manifold. Int. Math. Res. Not. IMRN (1999), Nr 3, 153–164

    1999

  34. [34]

    T. J. Lyons, Differential equations driven by rough signals. Revista Matem´ atica Iberoamericana, vol. 14(1998), Nr. 2, 215–310,1998, doi:10.4171/RMI/240

    work page doi:10.4171/rmi/240 1998

  35. [35]

    Finding the homology of submanifolds with high782 confidence from random samples.Discrete & Computational Geometry, 39(1-3):419–441, March 2008

    P. Niyogi, S. Smale, S. Weinberger, Finding the homology of submanifolds with high confidence from random samples. Discrete and Computational Geometry, vol. 39, nos. 1–3, 419-441 (2008)https://doi.org/10.1007/s00454-008-9053-2

    work page doi:10.1007/s00454-008-9053-2 2008

  36. [36]

    Rosenberg, The Laplacian on a Riemannian manifold : an introduction to anal- ysis on manifolds

    S. Rosenberg, The Laplacian on a Riemannian manifold : an introduction to anal- ysis on manifolds. London Mathematical Society student texts. Cambridge Uni- versity Press, Cambridge, U.K., New York, NY, USA, 1997

    1997

  37. [37]

    Singer, H.-T

    A. Singer, H.-T. Wu, Orientability and diffusion map. Appl. Comput. Harmon. Anal., 31(2011), 44–58

    2011

  38. [38]

    Singer, H.-T

    A. Singer, H.-T. Wu, Spectral convergence of the connection Laplacian from ran- dom samples. Information and Inference: A Journal of the IMA (2017) 6, 58–123 https://doi:10.1093/imaiai/iaw016

    work page doi:10.1093/imaiai/iaw016 2017

  39. [39]

    Singer, H.-T

    A. Singer, H.-T. Wu, Vector diffusion maps and the connection Laplacian. Comm. Pure Appl. Math., 65(2012), 1067–1144https://doi.org/10.1002/cpa.21395

    work page doi:10.1002/cpa.21395 2012

  40. [40]

    Talagrand, Sharper bounds for Gaussian and empirical processes

    M. Talagrand, Sharper bounds for Gaussian and empirical processes. Ann. Probab. 22 (1994) 28–76. EMPIRICAL HODGE LAPLACIANS AND MANIFOLD LEARNING 49

    1994

  41. [41]

    Talagrand, New concentration inequalities in product spaces

    M. Talagrand, New concentration inequalities in product spaces. Invent. Math. 126 (1996) 505–563

    1996

  42. [42]

    A. W. van der Vaart, J.A. Wellner, Weak convergence and Empirical Processes. 2nd Edition. Springer (1996)

    1996

  43. [43]

    Y. Yu, T. Wang, R.J. Samworth, A useful variant of the Davis–Kahan theorem for statisticians. Biometrika, 102(2015), N.2, 315-323. AppendixA.Proof of Proposition 4.2 The proof proceeds in three main steps: establishing a local coordinate representation, computing the eigengap of the population covariance opera- tor Σ t,δ(p), and applying concentration ine...

    2015