Pith. sign in

REVIEW 4 minor 26 references

Kirszbraun extensions preserving uniform distance in Hilbert spaces

T0 review · 0 major / 4 minor · reviewed 2026-08-01 · 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 a 1-Lipschitz extension with a uniform-distance constraint is exactly equivalent to a barycentric inequality, for arbitrary real Hilbert spaces, removing previous dimension and convexity restrictions.

desk verdict Ciosmak's equivalence is real: the new (E)⇒(C) direction is proven in full, and the imported converse from his earlier paper is legitimate. read the letter →

arxiv 2607.17672 v1 pith:VJSHALFW submitted 2026-07-20 math.FA math.MG

Krzysztof J. Ciosmak This is my paper · ORCID
classification math.FAmath.MG MSC 47H0954C2046C0549Q2260E15
keywords LipschitzextensionKirszbrauntheoremHilbertspacebarycentricinequalityuniformdistanceweaktransportdataprocessingconvexPoincaré
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

This paper proves that two conditions on a map v from a subset of a real Hilbert space into another real Hilbert space are exactly equivalent: (E) every 1-Lipschitz map on a subset that stays within a uniform distance ρ of v extends to a 1-Lipschitz map on the whole domain staying within ρ of v, and (C) a barycentric inequality saying that v never expands convex combinations more than the underlying points do. The new result is the implication from (E) to (C), previously known only for target spaces of dimension at most three or for convex domains. The proof introduces a minimal support simplex principle showing that any violation of (C) is witnessed by an affinely independent simplex whose scalar projection has a 1-Lipschitz affine interpolant, which then feeds into an isometric embedding argument. If correct, the equivalence resolves a conjecture for finite-dimensional targets, extends to infinite-dimensional Hilbert spaces, and yields new characterizations of data processing for a hierarchy of weak transport costs, plus transfer of convex Poincaré inequalities without loss of constant.

What carries the argument

The load-bearing object is the minimal support simplex principle (Proposition 2.1): in any real normed space, a strict scalar violation f(x0)-Σt_i f(x_i) > ||x0-Σt_i x_i|| admits a witness with affinely independent support and a 1-Lipschitz affine interpolant on that simplex. Corollary 2.2 calibrates this to vector-valued v by choosing w as the unit norming direction, giving a vector g with ||g||≤1 that linearly reproduces ⟨w,v(·)⟩ on the simplex. The proof of (E)⇒(C) then hinges on an explicit isometric linear embedding Q: L→Y satisfying Q^*w=g; that is possible because dim w⊥ ≥ dim L, and the inequality in Lemma 3.1 (the isometric test inequality) converts a candidate isometric copy into a

What would settle it

Exhibit a real Hilbert space pair Z,Y with dim Y ≥ 4, a subset X, and a map v:X→Y for which the barycentric condition (C) fails but the extension property (E) holds; or, to attack the converse direction, exhibit v satisfying (C) for which some 1-Lipschitz u within uniform distance ρ has no 1-Lipschitz extension within ρ of v. The four-point example in the paper satisfies (C) and is therefore not such a counterexample.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1.1: for real Hilbert spaces Z and Y, a map v:X→Y satisfies the extension property (E) if and only if it satisfies the barycentric condition (C), namely that for every k≤dim Y and every weighted average, ||v(x0)-Σt_i v(x_i)|| ≤ ||x0-Σt_i x_i||. The paper proves the previously open direction (E)⇒(C) by a limiting argument. Assuming (E) and a strict violation of (C), the minimal support simplex principle produces an affinely independent witness set together with a unit vector w and a vector g with ||g||≤1 that interpolates ⟨w,v(·)⟩ linearly on the witness simplex. Since the orthogonal complement of w has dimension at least the dimension of the simplex's linear span

Load-bearing premise

The full equivalence in Theorem 1.1 depends on the previously published theorem that the barycentric condition (C) suffices for the extension property (E); this paper proves only the reverse direction, so the equivalence as stated inherits that earlier result's correctness.

Editorial extensions

If this is right

  • For finite-dimensional targets, the equivalence resolves the conjecture on necessity of the barycentric condition, without requiring convexity of X or a bound on dim Y.
  • Condition (E) is characterized by the data processing inequality for the finite branching weak transport cost at branching level dim Y (Proposition 6.1), linking the extension property to contraction of transport costs.
  • When Y is infinite-dimensional or its dimension exceeds the affine dimension of X, property (E) implies the unrestricted barycentric condition (B), yielding a lifting theorem for convex Lipschitz functions (Proposition 6.2) and transfer of convex Poincaré inequalities with the same constant (Corollary 6.4).
  • The known four-point obstruction to the earlier proof method is bypassed: the proof only needs simplex witnesses, which always exist, so that example no longer blocks the characterization.

Reading between the lines

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

  • Because the minimal support simplex principle holds in arbitrary real normed spaces, a similar (E)⇔(C) characterization might extend to Banach spaces if the isometric embedding construction can be replaced by a weaker geometric argument; Hilbert-space structure enters mainly through the dimension count and the existence of Q.
  • The finite branch hierarchy B^[q] with q ≤ dim Y suggests a testable refinement: for a given v, the data processing inequality at level q may fail for all q smaller than dim Y even when (E) holds, indicating that the branching level is an essential parameter rather than a proof artifact.
  • The lifting theorem gives a constructive way to pull convex test functions back along v, so the convex Poincaré inequality transfer might be iterated along chains of maps preserving (E), offering a tool for proving concentration for composite Lipschitz maps.
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

0 major / 4 minor

Summary. The paper proves Theorem 1.1: for real Hilbert spaces Z, Y, any X⊂Z and v:X→Y, the uniform-distance-preserving Kirszbraun extension property (E) is equivalent to the barycentric condition (C). The new content is the implication (E)⇒(C), previously known only for dimY≤3 or convex X. The proof combines a minimal-support simplex principle (Prop. 2.1), a calibrated simplex witness (Cor. 2.2), an isometric test inequality (Lemma 3.1), and a rationalizing limit argument in §4. The converse (C)⇒(E) is quoted from the author's earlier paper [9, Theorem 1.2]. The paper also gives a finite-branching weak transport data-processing characterisation (Prop. 6.1) and, under a dimension assumption, a convex Lipschitz lifting theorem and transfer of convex Poincaré inequalities (Prop. 6.2, Cor. 6.4).

Significance. If correct, the main theorem resolves a conjecture from [7] and completes the equivalence of (E) and (C) for arbitrary real Hilbert targets, removing the dimension and convexity restrictions of earlier work. The new proof is self-contained: Proposition 2.1 is a clean independence result valid in arbitrary normed spaces, and the embedding Q in §4 is constructed explicitly. The applications to weak transport and convex Poincaré inequalities are natural and are derived without additional heavy machinery. The paper also explains the four-point obstruction from [9, Example 4.10], showing why the earlier method failed but the characterisation survives. The converse direction is imported from a published result of the author; this is disclosed and does not affect the novelty of the new implication.

minor comments (4)
  1. [§2, Proposition 2.1 proof] After choosing the minimizer p0, the display uses p* inconsistently; p* should be p0 throughout that paragraph. Please correct the notation.
  2. [Title/Abstract] The supplied version contains typesetting artifacts in the title and abstract (e.g., 'PRESER VING DIST ANCE' and stray glyphs in the displayed formula). Please ensure the final compiled version is clean.
  3. [§4] The construction of Q relies on the Riesz identification of L* with L; this is standard and correct, but a brief parenthetical 'after Riesz identification' would improve readability when stating Q*w = g.
  4. [§6.1] In the proof of Proposition 6.1, the equality σ(v(S),U') = σ(S,U) with U'=(S,U) is correct, but writing σ(v(S),S,U) would make the reasoning clearer and avoid the apparent omission of S.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the new (E)⇒(C) proof is self-contained; the sole self-cited converse is independently published and not used in the new derivation.

full rationale

Theorem 1.1's new direction, (E)⇒(C), is proved from scratch in Sections 2–4. Proposition 2.1 is a standalone extreme-point/affine-interpolation lemma valid in arbitrary normed spaces; Corollary 2.2 derives a calibrated vector witness; Lemma 3.1 converts an isometric embedding into a barycentric bound using (E); and the limiting rationalization argument yields the strict-barycentric contradiction. None of these steps assumes (C) or (E) as its conclusion. The only external ingredient in the overall equivalence is the converse (C)⇒(E), which the paper explicitly credits: "The reverse implication is proven in [9, Theorem 1.2]." This is the author's prior published theorem, not an input to the new proof, and its assumptions do not include the newly proved implication. The finite-branching data-processing and Poincaré-transfer applications are derived from Theorem 1.1 by direct two-way arguments, not by renaming. I find no equation that is equivalent to its input by construction and no fitted quantity relabelled as a prediction.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

No free parameters or fitted constants appear; R>0 in the limiting argument is a free variable, not fitted. The proof relies only on standard functional analysis and probability tools, all listed above. No new postulated entities are introduced; the finite-branching cost B[q] is a definition, not an additional axiom.

assumptions (5)
  • standard math Hahn–Banach supporting hyperplane theorem
    Used in the proof of Proposition 6.2 to represent a finite convex Lipschitz function f on Y as the supremum of its affine minorants.
  • standard math Riesz representation theorem in Hilbert spaces
    Used in Corollary 2.2 to identify the bounded linear functional ℓ on L with a vector g ∈ L of the same norm.
  • standard math Jensen's inequality for conditional expectations and convex functions
    Used in Proposition 6.1 to compute the finite-branching costs for one-point measures and to compare B[1] with the p-Wasserstein cost.
  • standard math Existence of regular conditional distributions on standard Borel spaces
    Invoked in Section 6.1 to justify the kernels in Γ_q and the pushforward construction in Proposition 6.1.
  • standard math Variance identity in Hilbert spaces (parallelogram law consequences)
    Used in Lemma 3.1 to convert pointwise length inequalities into an inequality for weighted barycenters.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Kirszbraun extensions preserving uniform distance in Hilbert spaces." pith.science (2026). https://pith.science/paper/VJSHALFW

@misc{pith2026260717672,
  author       = {Pith},
  title        = {Pith review of: Kirszbraun extensions preserving uniform distance in Hilbert spaces},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VJSHALFW}},
  note         = {Machine review of arXiv:2607.17672}
}
abstract

Let $X$ be a subset of a real Hilbert space and let $v\colon X\to Y$, where $Y$ is a real Hilbert space. We prove that the following conditions are equivalent: whenever $A\subset X$, $\rho\geq0$, and $u\colon A\to Y$ is $1$-Lipschitz with $\left\lVert u(x)-v(x)\right\lVert\leq\rho$ for $x\in A$, there is a $1$-Lipschitz extension $\widetilde u\colon X\to Y$ with $\left\lVert \widetilde u(x)-v(x)\right \lVert\leq\rho$ for $x\in X$; and for every $1\leq k\leq\dim Y$, $$ \left\lVert v(x_0)-\sum_{i=1}^k t_i v(x_i)\right\lVert \leq \left\lVert x_0-\sum_{i=1}^k t_i x_i \right \lVert$$ whenever $x_0,\ldots,x_k\in X$, $t_1,\ldots,t_k\geq0$, and $\sum_{i=1}^k t_i=1$. Previous necessity results required $\dim Y\leq3$ or convexity of $X$. For finite-dimensional targets, an application gives an exact data processing characterisation for a finite branching hierarchy connecting Wasserstein and barycentric weak transport. If $Y$ is infinite-dimensional or $\dim\operatorname{Aff}X+1\leq\dim Y$, we also obtain a lifting theorem for convex Lipschitz functions and transfer convex Poincar\'e inequalities without increasing the constant.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 19 canonical work pages

  1. [9]

    Ciosmak,Continuity of extensions of Lipschitz maps and of monotone maps, J

    K.J. Ciosmak,Continuity of extensions of Lipschitz maps and of monotone maps, J. Lond. Math. Soc. (2) 110 (5) (2024) e70014. doi:10.1112/jlms.70014

    work page doi:10.1112/jlms.70014 2024

  2. [7]

    Ciosmak,Optimal Transport and 1-Lipschitz Maps, DPhil thesis, University of Oxford, 2020

    K.J. Ciosmak,Optimal Transport and 1-Lipschitz Maps, DPhil thesis, University of Oxford, 2020. doi:10.5287/ora-nompqy5bn

    work page doi:10.5287/ora-nompqy5bn 2020

  3. [1]

    Adamczak, M

    R. Adamczak, M. Strzelecki,On the convex Poincaré inequality and weak transportation inequalities, Bernoulli 25 (1) (2019) 341–374. doi:10.3150/17-BEJ989

    work page doi:10.3150/17-bej989 2019

  4. [2]

    Azagra, E

    D. Azagra, E. Le Gruyer, C. Mudarra,Kirszbraun’s theorem via an explicit formula, Canad. Math. Bull. 64 (1) (2021) 142–153. doi:10.4153/S0008439520000314

    work page doi:10.4153/s0008439520000314 2021

  5. [3]

    Backhoff-Veraguas, M

    J. Backhoff-Veraguas, M. Beiglböck, G. Pammer,Existence, duality, and cyclical monotonicity for weak transport costs, Calc. Var. Partial Differential Equations 58 (2019), Paper No. 203, 28 pp. doi:10.1007/s00526-019-1624-y

    work page doi:10.1007/s00526-019-1624-y 2019

  6. [4]

    Backhoff-Veraguas, G

    J. Backhoff-Veraguas, G. Pammer,Applications of weak transport theory, Bernoulli 28 (1) (2022) 370–394. doi:10.3150/21-BEJ1346

    work page doi:10.3150/21-bej1346 2022

  7. [5]

    Beiglböck, G

    M. Beiglböck, G. Pammer, L. Riess, S. Schrott,The fundamental theorem of weak optimal transport, preprint (2025). arXiv:2501.16316

    arXiv 2025

  8. [6]

    Cavagnari, G

    G. Cavagnari, G. Savaré, G.E. Sodini,Extension of monotone operators and Lipschitz maps invariant for a group of isometries, Canad. J. Math. 77 (1) (2025) 149–186. doi:10.4153/S0008414X23000846

    work page doi:10.4153/s0008414x23000846 2025

Show all 26 references
  1. [8]

    Ciosmak,Continuity of extensions of Lipschitz maps, Israel J

    K.J. Ciosmak,Continuity of extensions of Lipschitz maps, Israel J. Math. 245 (2) (2021) 567–588. doi:10.1007/s11856-021-2215-0

    2021 doi

  2. [10]

    Espínola, A

    R. Espínola, A. Nicolae,Continuous selections of Lipschitz extensions in metric spaces, Rev. Mat. Complut. 28 (3) (2015) 741–759. doi:10.1007/s13163-015-0171-0

    2015 doi

  3. [11]

    Feldheim, A

    N. Feldheim, A. Marsiglietti, P. Nayar, J. Wang,A note on the convex infimum convolution inequality, Bernoulli 24 (1) (2018) 257–270. doi:10.3150/16-BEJ875

    2018 doi

  4. [12]

    Gozlan, N

    N. Gozlan, N. Juillet,On a mixture of Brenier and Strassen theorems, Proc. Lond. Math. Soc. (3) 120 (3) (2020) 434–463. doi:10.1112/plms.12302

    2020 doi

  5. [13]

    Gozlan, C

    N. Gozlan, C. Roberto, P.-M. Samson, Y. Shu, P. Tetali,Characterization of a class of weak transport-entropy inequalities on the line, Ann. Inst. Henri Poincaré Probab. Stat. 54 (3) (2018) 1667–1693. doi:10.1214/17-AIHP851

    2018 doi

  6. [14]

    Gozlan, C

    N. Gozlan, C. Roberto, P.-M. Samson, P. Tetali,Kantorovich duality for gen- eral transport costs and applications, J. Funct. Anal. 273 (11) (2017) 3327–3405. doi:10.1016/j.jfa.2017.08.015

    2017 doi

  7. [15]

    Kallenberg,Foundations of Modern Probability, 3rd ed., Probability Theory and Stochastic Modelling, vol

    O. Kallenberg,Foundations of Modern Probability, 3rd ed., Probability Theory and Stochastic Modelling, vol. 99, Springer, Cham, 2021. doi:10.1007/978-3-030-61871-1

    2021 doi

  8. [16]

    Kirszbraun,Über die zusammenziehende und Lipschitzsche Transformationen, Fund

    M. Kirszbraun,Über die zusammenziehende und Lipschitzsche Transformationen, Fund. Math. 22 (1934) 77–108. doi:10.4064/fm-22-1-77-108

    1934 doi

  9. [17]

    Kopecká,Bootstrapping Kirszbraun’s extension theorem, Fund

    E. Kopecká,Bootstrapping Kirszbraun’s extension theorem, Fund. Math. 217 (1) (2012) 13–19. doi:10.4064/fm217-1-2

    2012 doi

  10. [18]

    Kopecká,Extending Lipschitz mappings continuously, J

    E. Kopecká,Extending Lipschitz mappings continuously, J. Appl. Anal. 18 (2) (2012) 167–177. doi:10.1515/jaa-2012-0011. KIRSZBRAUN EXTENSIONS PRESER VING UNIFORM DISTANCE 17

    2012 doi

  11. [19]

    Kopecká, S

    E. Kopecká, S. Reich,Continuous extension operators and convexity, Nonlinear Anal. 74 (18) (2011) 6907–6910. doi:10.1016/j.na.2011.07.013

    2011 doi

  12. [20]

    U. Lang, V. Schroeder,Kirszbraun’s theorem and metric spaces of bounded curvature, Geom. Funct. Anal. 7 (3) (1997) 535–560. doi:10.1007/s000390050018

    1997 doi

  13. [21]

    Marton,Bounding ¯d-distance by informational divergence: a method to prove meas- ure concentration, Ann

    K. Marton,Bounding ¯d-distance by informational divergence: a method to prove meas- ure concentration, Ann. Probab. 24 (2) (1996) 857–866. doi:10.1214/aop/1039639365

    1996

  14. [22]

    Reich, S

    S. Reich, S. Simons,Fenchel duality, Fitzpatrick functions and the Kirszbraun– Valentine extension theorem, Proc. Amer. Math. Soc. 133 (9) (2005) 2657–2660. doi:10.1090/S0002-9939-05-07983-9

    2005 doi

  15. [23]

    Y. Shu, M. Strzelecki,A characterization of a class of convex log-Sobolev inequalities on the real line, Ann. Inst. Henri Poincaré Probab. Stat. 54 (4) (2018) 2075–2091. doi:10.1214/17-AIHP865

    2018 doi

  16. [24]

    Talagrand,Concentration of measure and isoperimetric inequalities in product spaces, Publ

    M. Talagrand,Concentration of measure and isoperimetric inequalities in product spaces, Publ. Math. Inst. Hautes Ètudes Sci. 81 (1995) 73–205. doi:10.1007/BF02699376

    1995 doi

  17. [25]

    Valentine,A Lipschitz condition preserving extension for a vector function, Amer

    F.A. Valentine,A Lipschitz condition preserving extension for a vector function, Amer. J. Math. 67 (1945) 83–93. doi:10.2307/2371917

    1945 doi

  18. [26]

    Wintenberger,Weak transport inequalities and applications to exponential and oracle inequalities, Electron

    O. Wintenberger,Weak transport inequalities and applications to exponential and oracle inequalities, Electron. J. Probab. 20 (2015), Paper No. 114, 27 pp. doi:10.1214/EJP.v20- 3558. Beijing Institute of Mathematical Sciences and Applications, No. 544, Hefangkou Village, Huaibe...

    2015 doi

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.