pith. sign in

arxiv: 2604.21340 · v1 · submitted 2026-04-23 · 🧮 math.NA · cs.NA· math.NT

Spherical Cap L₂ Discrepancy -- Blessing of Dimensionality and a Balanced Large-Cap Variant

Johann S. Brauchart , Josef Dick , Friedrich Pillichshammer This is my paper

Pith reviewed 2026-05-09 21:30 UTC · model grok-4.3

classification 🧮 math.NA cs.NAmath.NT
keywords spherical cap discrepancyL2 discrepancyinformation complexityStolarsky invarianceSobolev spacenumerical integrationsphereblessing of dimensionality
0
0 comments X

The pith

The information complexity of classical spherical cap L2 discrepancy on the d-sphere decreases with increasing dimension.

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

The paper establishes that the classical spherical cap L2 discrepancy on the d-dimensional sphere has information complexity that decreases as d increases, so fewer points suffice to reach a given discrepancy level in higher dimensions. This points to a blessing of dimensionality for numerical integration tasks that rely on such discrepancies. The authors then define a modified version that weights large caps more heavily. For this variant, Stolarsky invariance equates the discrepancy to the worst-case integration error in the Sobolev space H^{(d+1)/2} on the sphere, and that error grows polynomially in d.

Core claim

We prove that the information complexity of the classical spherical cap L2 discrepancy on S^d decreases with d. We introduce a modified spherical cap L2 discrepancy that emphasizes large caps close to hemispheres. For this variant we establish a Stolarsky invariance principle that connects the discrepancy to numerical integration in the Sobolev space H^{(d+1)/2}(S^d) with reproducing kernel K(x,y)=1-1/sqrt(2)||x-y||, and this connection implies that the worst-case integration error grows polynomially with d.

What carries the argument

The Stolarsky invariance principle, which equates the modified spherical cap L2 discrepancy directly to the worst-case integration error in H^{(d+1)/2}(S^d) with the kernel 1 - 1/sqrt(2) ||x-y||.

If this is right

  • Fewer points are needed to achieve a fixed L2 discrepancy level on the sphere as dimension grows for the classical definition.
  • The modified large-cap version yields a discrepancy measure whose associated integration error increases polynomially with d.
  • Numerical integration problems on the sphere using the classical discrepancy become easier in high dimensions while the modified version remains stable or harder.
  • Point distributions optimized for the modified discrepancy do not gain efficiency from higher dimensions.

Where Pith is reading between the lines

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

  • The contrast between the two variants suggests that the choice of cap weighting can control whether dimensionality helps or hinders discrepancy-based integration.
  • The polynomial growth result may guide the design of point sets when large hemispherical regions matter more than small caps.

Load-bearing premise

The Stolarsky invariance principle applies exactly to the modified spherical cap L2 discrepancy and equates it to the worst-case error in the Sobolev space H^{(d+1)/2}(S^d).

What would settle it

A numerical computation of the minimal spherical cap L2 discrepancy for increasing d that shows the information complexity does not decrease, or an explicit counterexample demonstrating that Stolarsky invariance fails for the modified large-cap version.

read the original abstract

We prove that the information complexity (i.e., the inverse) of the classical spherical cap $L_2$ discrepancy on the $d$-dimensional sphere $\mathbb{S}^d$ decreases with dimension $d$, indicating a ``blessing of dimensionality'' for the associated numerical integration problem. We then introduce a modified spherical cap $L_2$ discrepancy that emphasizes large caps (close to hemispheres). For this variant, the problem does not become easier with increasing $d$. We also establish a Stolarsky invariance principle which connects the modified spherical cap $L_2$ discrepancy to numerical integration in the Sobolev space $H^{(d+1)/2}(\mathbb{S}^d)$, represented by the reproducing kernel $K(\boldsymbol{x}, \boldsymbol{y}) = 1 - \tfrac{1}{\sqrt{2}} \|\boldsymbol{x} - \boldsymbol{y}\|$. Stolarsky's invariance principle then implies that the worst-case integration error in this space grows polynomially with $d$.

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

2 major / 1 minor

Summary. The paper proves that the information complexity of the classical spherical cap L2 discrepancy on S^d decreases with dimension d. It introduces a modified large-cap variant of the discrepancy and establishes a Stolarsky invariance principle equating this variant to the squared worst-case integration error in the Sobolev space H^{(d+1)/2}(S^d) with reproducing kernel K(x,y)=1-(1/sqrt(2))||x-y||, from which it concludes that the worst-case error grows polynomially in d.

Significance. If the derivations are correct, the results provide a clear demonstration of dimensional blessing for the classical discrepancy and a dimensionally robust lower bound for the modified variant via an explicit Sobolev-space equivalence. The Stolarsky link supplies a concrete, kernel-based characterization that could be useful for quadrature error analysis on spheres.

major comments (2)
  1. [Abstract and Stolarsky-invariance section] The identification of K(x,y)=1-(1/sqrt(2))||x-y|| as the reproducing kernel of H^{(d+1)/2}(S^d) is load-bearing for the polynomial-growth claim. The standard RK for this space has Gegenbauer coefficients proportional to [l(l+d-1)]^{-(d+1)/2} (normalized by surface measure); the manuscript must verify that the chordal-distance kernel possesses exactly these multipliers for every d, including the precise normalization factor.
  2. [Classical-discrepancy section] The proof that the information complexity of the classical spherical-cap L2 discrepancy decreases with d is stated but not inspectable in the provided text. The explicit upper bound on the minimal number of points N(d,ε) (or the rate at which it tends to zero) should be displayed, together with the argument that the constant factors remain controlled as d→∞.
minor comments (1)
  1. [Introduction of modified discrepancy] Notation for the modified discrepancy functional should be introduced with an explicit integral formula (rather than only by verbal description) to make the large-cap weighting unambiguous.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and constructive comments. We address each major point below and will revise the manuscript to improve inspectability and completeness.

read point-by-point responses

  1. Referee: [Abstract and Stolarsky-invariance section] The identification of K(x,y)=1-(1/sqrt(2))||x-y|| as the reproducing kernel of H^{(d+1)/2}(S^d) is load-bearing for the polynomial-growth claim. The standard RK for this space has Gegenbauer coefficients proportional to [l(l+d-1)]^{-(d+1)/2} (normalized by surface measure); the manuscript must verify that the chordal-distance kernel possesses exactly these multipliers for every d, including the precise normalization factor.

    Authors: We agree that an explicit verification strengthens the Stolarsky-invariance argument. In the revised manuscript we will add a dedicated calculation (in the Stolarsky section) expanding the chordal kernel in Gegenbauer polynomials and confirming that its coefficients are exactly proportional to [l(l+d-1)]^{-(d+1)/2} with the normalization factor independent of d. This will directly support the polynomial-growth claim for the worst-case error. revision: yes

  2. Referee: [Classical-discrepancy section] The proof that the information complexity of the classical spherical-cap L2 discrepancy decreases with d is stated but not inspectable in the provided text. The explicit upper bound on the minimal number of points N(d,ε) (or the rate at which it tends to zero) should be displayed, together with the argument that the constant factors remain controlled as d→∞.

    Authors: We apologize for the insufficient detail in the excerpt. The proof proceeds by deriving an explicit upper bound on the minimal N(d,ε) that tends to zero with d (while the discrepancy is held below ε). In the revision we will display this bound together with the accompanying estimates, explicitly showing that all constants remain controlled (in fact improve) as d→∞. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation relies on independent proofs and established invariance

full rationale

The paper proves bounds showing information complexity of classical spherical cap L2 discrepancy decreases with d, introduces a modified large-cap variant, establishes a Stolarsky invariance principle equating the modified discrepancy to worst-case error in H^{(d+1)/2}(S^d) via the explicit kernel K(x,y)=1-(1/sqrt(2))||x-y||, and concludes polynomial growth in d for the error. All steps are self-contained mathematical arguments with no reduction by construction to fitted parameters, self-definitions, or load-bearing self-citations. The invariance is derived in the paper rather than assumed from prior author work, and the kernel identification is presented as part of the established connection without tautological renaming or smuggling. This is the normal case of a theoretical paper whose central claims have independent content.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The work rests on standard definitions of spherical caps, L2 discrepancy, and the Stolarsky principle; no free parameters are fitted and no new entities are postulated beyond the modified discrepancy definition itself.

axioms (2)
  • standard math Standard mathematical properties of spherical caps, L2 norms, and information complexity on the sphere S^d
    Used to define the classical and modified discrepancy measures.
  • domain assumption The Stolarsky invariance principle holds and equates the modified discrepancy to the worst-case error in the Sobolev space with kernel K(x,y) = 1 - 1/sqrt(2) ||x-y||
    Central to linking the discrepancy to integration error and deriving the polynomial growth.

pith-pipeline@v0.9.0 · 5492 in / 1380 out tokens · 48440 ms · 2026-05-09T21:30:12.968010+00:00 · methodology

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

15 extracted references · 15 canonical work pages

  1. [1]

    Bilyk, J

    D. Bilyk, J. S. Brauchart: On the lower bounds for the spherical cap discrepancy. arXiv:2502.15984

    work page arXiv

  2. [2]

    Bilyk, F

    D. Bilyk, F. Dai: Geodesic distance Riesz energy on the sphere. Trans. Amer. Math. Soc. 372(5): 3141–3166, 2019

    work page 2019

  3. [3]

    Bilyk, F

    D. Bilyk, F. Dai, R. Matzke: The Stolarsky principle and energy optimization on the sphere. Constr. Approx. 48(1): 31–60, 2018

    work page 2018

  4. [4]

    J. S. Brauchart, J. Dick: A simple proof of Stolarsky’s invariance principle. Proc. Amer. Math. Soc. 141(6): 2085–2096, 2013

    work page 2085

  5. [5]

    J. S. Brauchart, J. Dick: A characterization of Sobolev spaces on the sphere and an extension of Stolarsky’s invariance principle to arbitrary smoothness. Constr. Approx. 38(3): 397–445, 2013

    work page 2013

  6. [6]

    J. S. Brauchart, E. B. Saff, I. H. Sloan, R. S. Womersley: QMC designs: optimal order quasi Monte Carlo integration schemes on the sphere. Math. Comp. 83(290): 2821–2851, 2014

    work page 2014

  7. [7]

    Choirat, R

    C. Choirat, R. Seri: Numerical properties of generalized discrepancies on spheres of arbitrary dimension, J. Complexity 29(2): 216–235, 2013

    work page 2013

  8. [8]

    J. Dick, F. Pillichshammer: Discrepancy theory and quasi-Monte Carlo integration. In:A Panorama of Discrepancy Theory(W. Chen, A. Srivastav, G. Travaglini, eds.), pp. 539–619, Lecture Notes in Math., No. 2107, Springer, Cham, 2014

    work page 2014

  9. [9]

    M¨ uller:Spherical Harmonics.Lecture Notes in Math., No

    C. M¨ uller:Spherical Harmonics.Lecture Notes in Math., No. 17, Springer, Berlin, 1966

    work page 1966

  10. [10]

    NIST Digital Library of Mathematical Functions.https://dlmf.nist.gov/, Release 1.2.4 of 2025-03-15. F. W. J. Olver, A. B. Olde Daalhuis, D. W. Lozier, B. I. Schneider, R. F. Boisvert, C. W. Clark, B. R. Miller, B. V. Saunders, H. S. Cohl, and M. A. McClain, eds

    work page 2025

  11. [11]

    Novak, F

    E. Novak, F. Pillichshammer: Intractability results for integration in tensor product spaces. J. Complexity 86: Paper No. 101901, 28 pp., 2025

    work page 2025

  12. [12]

    Novak, F

    E. Novak, F. Pillichshammer: TheL p-discrepancy for finitep >1 suffers from the curse of dimensionality. Proc. Amer. Math. Soc. 153(4): 1447–1459, 2025

    work page 2025

  13. [13]

    Novak, H

    E. Novak, H. Wo´ zniakowski:Tractability of Multivariate Problems. Volume I: Linear Information.EMS Tracts in Mathematics 16, Z¨ urich, 2008

    work page 2008

  14. [14]

    Novak, H

    E. Novak, H. Wo´ zniakowski:Tractability of Multivariate Problems. Volume II: Stan- dard Information for Functionals.EMS Tracts in Mathematics 12, Z¨ urich, 2010

    work page 2010

  15. [15]

    Wo´ zniakowski: Efficiency of quasi-Monte Carlo algorithms for high dimensional integrals

    H. Wo´ zniakowski: Efficiency of quasi-Monte Carlo algorithms for high dimensional integrals. In:Monte Carlo and Quasi-Monte Carlo Methods 1998(H. Niederreiter and J. Spanier, eds.), pp. 114–136, Springer Verlag, Berlin, 2000. 17

    work page 1998