Pith. sign in

REVIEW 1 major objections 5 minor 1 cited by

The fractional Helly number for separable convexity spaces

T0 review · 1 major / 5 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 fractional Helly number of a separable convexity space is bounded by the dual VC-dimension of its halfspaces plus one.

desk verdict Main theorem is solid and worth citing; the Bárány–Kalai disproof has a gap (B_2^2 is not a convexity space), but that is a side result and the central argument holds up. read the letter →

arxiv 2412.01445 v4 pith:JC6P5TAY submitted 2024-12-02 math.CO

Andreas F. Holmsen , Zuzana Patáková This is my paper · ORCID
classification math.CO MSC 52A3552A01
keywords fractionalHellynumberconvexityspacesRadondualVC-dimensionhalfspacesseparableconvexlatticesetsbox
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

Separable convexity spaces are abstract set systems with a notion of convex set that satisfies closure axioms and a separation axiom: any convex set and an exterior point can be separated by a halfspace, where a halfspace is a convex set whose complement is also convex. This paper proves that if the halfspaces of such a space have dual VC-dimension $d$, then the fractional Helly number of the whole family of convex sets is at most $d+1$. Consequently, in any separable convexity space with Radon number at most $r$, the fractional Helly number is at most $2^r$. This recovers the Bárány–Matoušek theorem for convex lattice sets in $\mathbb{Z}^d$, and the bound is asymptotically tight, as shown by box convexity in $\mathbb{R}^d$. The same framework also disproves a conjecture of Bárány and Kalai about fractional Helly properties for solutions of bounded-degree polynomial inequalities.

What carries the argument

The load-bearing object is Proposition 8, a weak colorful Helly theorem for separable convexity spaces with bounded Radon number. It says that if $d+1$ families of convex sets, each of size $p$, have the property that every transversal choice of one set from each family has nonempty intersection, then one of the families contains $m$ members with nonempty intersection. The proof of Proposition 8 relies on Lemma 10, which uses the separability axiom to produce a halfspace separating one convex set from an exterior point, and Lemma 11, which applies the Radon bound and the Erdős–Simonovits supersaturation theorem to find enough separable vertices. The argument concludes by forcing $d+1$ halfspaces to realize a complete Venn diagram, contradicting the assumption that the dual VC-dimension is at most $d$.

What would settle it

A single example would refute the main theorem: a separable convexity space with Radon number at most $r$ whose fractional Helly number exceeds $2^r$, or with halfspaces of dual VC-dimension $d$ whose fractional Helly number is $d+2$. A more accessible check is the box-convexity example itself, which the paper claims has fractional Helly number exactly $d+1$; constructing a family of axis-parallel boxes with an $\alpha$-fraction of intersecting $d$-tuples but no large intersecting subfamily for $\alpha$ arbitrarily close to $1$ would falsify that example's tightness.

Watch

Extended reading notes

Core claim

The central claim is Theorem 5: for a separable convexity space $(X,\mathcal{C})$ with bounded Radon number, if the system of halfspaces has dual VC-dimension $d$, then the fractional Helly number for $\mathcal{C}$ is at most $d+1$. Corollary 6 then yields the exponential bound $2^r$ in terms of the Radon number $r$. The paper shows the exponential bound cannot be improved asymptotically: for box convexity on $\mathbb{R}^d$, the Radon number is $\Theta(\log d)$ and the fractional Helly number equals $d+1$. It also proves that the convexity space $B^2_2$ of solutions of polynomial inequalities of degree at most two in $\mathbb{R}^2$ is universal for intersection patterns, making its Helly, Radon, and fractional Helly numbers all unbounded, which disproves Conjecture 2.9 of Bárány and Kalai.

Load-bearing premise

The argument assumes the separability axiom S3, which guarantees that any convex set and any point outside it can be separated by a halfspace; if this axiom fails, the proof has no way to produce the separating halfspace that the argument needs.

Editorial extensions

If this is right

  • The fractional Helly number of convex lattice sets in $\mathbb{Z}^d$ is $d+1$, recovering the Bárány–Matoušek theorem as a special case of Theorem 5.
  • In every separable convexity space with Radon number at most $r$, the fractional Helly number is at most $2^r$, and box convexity shows that this exponential dependence is asymptotically unavoidable.
  • The Bárány–Kalai conjecture on fractional Helly properties for bounded-degree polynomial inequalities is false: already in $B^2_2$ the Helly, Radon, and fractional Helly numbers are all unbounded.
  • If a separable convexity space has halfspaces with VC-dimension at most $d$, then the fractional Helly number is bounded by a function of $d$, giving an affirmative answer to a problem of Bárány and Kalai once the bounded-Radon assumption is added.

Reading between the lines

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

  • Since the dual VC-dimension is used only in the final step of the proof, one might expect the same bound $d+1$ to hold for any separable convexity space whose halfspaces cannot shatter $d+1$ sets in the dual sense, even if the halfspace system is not literally a set system of bounded dual VC-dimension.
  • The proof suggests that separability, rather than any metric or lattice structure, is the property that lets local Radon bounds become global fractional Helly bounds; testing weaker separation axioms (for example, separation only for finite convex sets) would delineate the exact boundary.
  • The universality of $B^2_2$ implies that any convexity space that contains $B^2_2$ as a subspace should also have unbounded Helly parameters; this may guide searches for other 'wild' convexity spaces.
  • A direct consequence one can test: in any separable convexity space with Radon number $r$, the fractional Helly number should be exactly $d+1$ for the minimal dual VC-dimension $d$ of its halfspaces; verifying this for concrete spaces like box convexity in higher dimensions would confirm the tightness beyond the asymptotic statement.
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

1 major / 5 minor

Summary. The paper proves that in a separable convexity space with bounded Radon number, if the dual VC-dimension of the halfspace system is d, then the fractional Helly number of the convexity space is at most d+1. As a corollary, the fractional Helly number is at most 2^r when the Radon number is r. This generalizes the Bárány–Matoušek theorem for convex lattice sets. The paper also claims to disprove a conjecture of Bárány and Kalai by showing that the set system of solutions of bounded-degree polynomial inequalities is universal.

Significance. The main theorem is a significant and clean generalization of the Bárány–Matoušek result, with a self-contained proof of the key colorful Helly proposition. The 2^r bound is near-optimal, as illustrated by box convexity. The proof uses Holmsen–Lee only as an independent, weaker result, and it contains no circularity or fitted parameters. However, the supplementary disproof of the Bárány–Kalai conjecture is compromised by an error in the claim that B_d^k is a convexity space.

major comments (1)
  1. [1.8, Proposition 7] The claim that B_d^k is a separable convexity space is false because the family of solution sets of finitely many degree-≤k polynomial inequalities is not closed under nested unions, violating axiom (C3). For d=k=2, the construction inside Proposition 7 realizes every finite subset of the x-axis as an element of B_2^2: for any a_1 < ... < a_{n+1}, the system {a_1 ≤ x ≤ a_{n+1}} ∪ {0 ≤ y ≤ (x−a_i)(x−a_{i+1})}_{i=1}^n has exactly the points (a_i,0) as its solution set. Hence the sets S_n = {(1,0),...,(n,0)} all lie in B_2^2 and form a chain under inclusion. Their union N×{0} is not a solution set of a finite polynomial system, since every finite semialgebraic set has finitely many connected components while N×{0} has infinitely many. Thus B_2^2 violates (C3). Consequently, the Radon, Helly, and fractional Helly numbers of B_2^2 are not defined, and the claimed disproof of Conjecture 2.9 does not follow from Proposition 7 as stated. The authors should either construct a genuine convexity space that still has the universality property, or rephrase the disproof directly for the set system of polynomial solution sets, which does not require the convexity-space axioms.
minor comments (5)
  1. [4, proof of Proposition 8] The sentence 'this can be repeated another d times' is too terse. The authors should explain how the already-fixed 2-element vertex classes are used when Lemma 11 is applied to the remaining classes, and why the separability of a fixed pair is preserved when the other classes are later shrunk.
  2. [Statement of Proposition 8] The notation '⋂s i=1' in the hypothesis should read '⋂_{i=1}^{d+1}'.
  3. [Proof of Lemma 11] There is a typo: 'there will exists s-element subsets' should be 'there will exist s-element subsets'.
  4. [1.8] The phrase 'B_d^k is in a sense "universal"' could be clarified: Proposition 7 demonstrates universality for intersection patterns of set systems, not for all convexity-theoretic notions.
  5. [1.2] The definition of the fractional Helly property uses 'the fractional Helly number for C'; for consistency with the rest of the paper, consider using 'of C'.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the proof of Theorem 5 is self-contained, and the Holmsen–Lee theorem is used as an independent weaker external result rather than as the target conclusion.

full rationale

The derivation chain for Theorem 5 is not circular. The proof starts from a family of convex sets with many intersecting (d+1)-tuples, builds the corresponding (d+1)-uniform hypergraph, applies Erdős–Simonovits supersaturation to obtain many copies of K_{d+1}(p), and feeds each copy into Proposition 8. Proposition 8 is proved in Section 3 using only the separation axiom, bounded Radon number, and the definition of dual VC-dimension; the final contradiction is that d+1 halfspaces would realize a complete Venn diagram, directly contradicting the assumed dual VC-dimension bound. The only author-overlapping citation is Theorem 3 (Holmsen–Lee), which is used at the very end to convert many intersecting m-tuples into a large intersecting subfamily. That theorem is an independent, published, weaker result: it already establishes some fractional Helly property depending only on the Radon number, and the present paper strengthens the number to d+1. This is a legitimate black-box use, not an assumption of the target theorem. There are no fitted parameters, no definitional equivalences, and no renaming of a known result. The notable gap concerning whether B_2^2 satisfies axiom (C3) in Section 1.8 is a mathematical correctness issue, not a circularity issue, and does not affect the main theorem's derivation.

Assumptions & free parameters 0 free parameters · 9 assumptions · 0 invented entities

The central theorem introduces no free parameters and no new entities. It relies on standard external theorems from convexity, VC theory, and extremal hypergraph theory, plus the structural assumptions of separability, bounded Radon number, and bounded dual VC-dimension. The only ad hoc input is the assertion, used for the counterexample, that B_d^k forms a separable convexity space.

assumptions (9)
  • standard math Levi's theorem: Radon number r implies Helly number at most r−1
    Used in Lemma 10 to conclude the family F of convex hulls has a common point.
  • standard math Erdős–Simonovits supersaturation theorem for hypergraphs
    Used in Lemma 11 and in the proof of Theorem 5 to extract complete multipartite subhypergraphs from dense hypergraphs.
  • standard math Assouad's inequality: dual VC-dim(F) < 2^{VC-dim(F)+1}
    Used in Corollary 6 to pass from halfspace VC-dimension at most r−1 to dual VC-dimension at most 2^r.
  • standard math Moran–Yehudayoff lemma: halfspaces of a separable convexity space with Radon number r have VC-dimension at most r−1
    Used in Corollary 6 to bound the VC-dimension of halfspaces.
  • standard math Holmsen–Lee theorem: any convexity space with bounded Radon number has the fractional Helly property
    Used at the end of the proof of Theorem 5 to convert many intersecting m-tuples into a large intersecting subfamily.
  • domain assumption Separability (S3): every convex set and exterior point are separated by a halfspace
    The theorem assumes it and Lemma 10 requires it to construct the separating halfspace.
  • domain assumption Convexity space axioms (C1)-(C3): empty set, X, closed under intersections and nested unions
    The entire framework of the paper assumes these axioms.
  • domain assumption The system of halfspaces has finite dual VC-dimension d
    Theorem 5's hypothesis; used in the final Venn diagram contradiction.
  • ad hoc to paper B_d^k is a separable convexity space
    Asserted without proof; closure under nested unions is not established for bounded-degree polynomial inequality sets and may fail. Placed as ad hoc because it supports only the counterexample, not the main theorem.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The fractional Helly number for separable convexity spaces." pith.science (2026). https://pith.science/paper/JC6P5TAY

@misc{pith2026241201445,
  author       = {Pith},
  title        = {Pith review of: The fractional Helly number for separable convexity spaces},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JC6P5TAY}},
  note         = {Machine review of arXiv:2412.01445}
}
abstract

A convex lattice set in $\mathbb{Z}^d$ is the intersection of a convex set in $\mathbb{R}^d$ with the integer lattice $\mathbb{Z}^d$. A classical theorem of Doignon states that the Helly number of $d$-dimensional convex lattice sets equals $2^d$, exponentially larger than the Helly number $d+1$ of ordinary convex sets in $\mathbb{R}^d$. By contrast, a remarkable theorem of B\'ar\'any and Matousek states that the fractional Helly number of convex lattice sets drops back down to $d+1$, matching the classical fractional Helly theorem of Katchalski and Liu. In this paper we generalize the B\'ar\'any--Matousek theorem to abstract convexity spaces (in the sense of van de Vel) that satisfy a suitable separation axiom. Our main result implies the following: if a separable convexity space has Radon number at most $r$, then its fractional Helly number is at most $2^{r}$. This bound is nearly tight, as illustrated by the case of box convexity in $\mathbb{R}^d$, whose Radon number is $\Theta(\log d)$ and fractional Helly number equals $d+1$.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. Helly-type theorems for separated $d$-intervals

    math.CO 2025-01 reject novelty 6.0 of 10

    The paper asserts that nerves of separated d-interval families are (2d-1)-collapsible, yielding Helly-type theorems for the associated convexity spaces.

Reference graph

Works this paper leans on

21 extracted references · 21 canonical work pages · cited by 1 Pith paper

  1. [1]

    N. Alon, G. Kalai, J. Matoušek, and R. Meshulam. Transver sal numbers for hypergraphs arising in geometry. Adv. in Appl. Math. 29 (2002), 79–101

    work page 2002

  2. [2]

    Alon and D

    N. Alon and D. J. Kleitman. Piercing convex sets and Hadwi ger–Debrunner (p,q) problem. Adv. Math. 96 (1992), 103–112

    work page 1992

  3. [3]

    P . Assouad. Densité et dimension. Ann. Inst. Fourier 33 ( 1983), 233–282

    work page 1983

  4. [4]

    I. Bárány. Combinatorial Convexity. Amer. Math. Soc., University Lecture Series 77, 2021

    work page 2021

  5. [5]

    Bárány and G

    I. Bárány and G. Kalai. Helly-type problems. Bull. Amer. Math. Soc. 59 (2022), 471–502. THE FRACTIONAL HELLY NUMBER FOR SEPARABLE CONVEXITY SPACES 11

    work page 2022

  6. [6]

    Bárány and J

    I. Bárány and J. Matoušek. A fractional Helly theorem for convex lattice sets. Adv. Math. 174 (2003), 227–235

    work page 2003

  7. [7]

    J. P . Doignon. Convexity in cristallographical lattice s. J. Geom. 3 (1973), 71–85

    work page 1973

  8. [8]

    J. Eckho ff. An upper-bound theorem for families of convex sets. Geom. D edicata 19 (1985), 217–227

    work page 1985

Show all 21 references
  1. [9]

    J. Eckho ff. The partition conjecture. Discrete Math. 221 (2000), 61–7 8

    2000

  2. [10]

    Erd˝os and M

    P . Erd˝os and M. Simonovits. Supersaturated graphs and hypergraph s. Combinatorica 3 (1983), 181–192

    1983

  3. [11]

    E. Helly. Über mengen konvexer körper mit gemeinschaft lichen punkte. Jahresber. Deutsch. Math.-V erein. 32 (1923), 175–176

    1923

  4. [12]

    A. F. Holmsen and D. Lee. Radon numbers and the fractiona l Helly theorem. Israel J. Math. 241 (2021), 433–447

    2021

  5. [13]

    G. Kalai. Intersection patterns of convex sets. Israel J. Math. 48 (1984), 175–195

    1984

  6. [14]

    Katchalski and A

    M. Katchalski and A. Liu. A problem of geometry in Rn. Proc. Amer. Math. Soc. 75 (1979), 284–288

    1979

  7. [15]

    F. W . Levi. On Helly’s theorem and the axioms of convexit y. J. Indian Math. Soc. 15 (1951), 65–76

    1951

  8. [16]

    Martinez-Legaz and I

    J.-E. Martinez-Legaz and I. Singer. The structure of he mispaces in Rn. Linear Algebra Appl. 110 (1998), 117–179

    1998

  9. [17]

    Matoušek

    J. Matoušek. Lectures on Discrete Geometry, Springer GTM 212, 2002

    2002

  10. [18]

    Matoušek

    J. Matoušek. Bounded VC-dimension implies a fractiona l Helly theorem. Discrete Com- put. Geom. 31 (2004), 251–255

    2004

  11. [19]

    Moran and A

    S. Moran and A. Y ehudayoff. On weak ε-nets and the Radon number. Discrete Comput. Geom. 64 (2020), 1125–1140

    2020

  12. [20]

    S. Onn. On the geometry and computational complexity of Radon partitions in the integer lattice. SIAM J. Discrete Math. 4 (1991), 436–446

    1991

  13. [21]

    M. L. J. van de V el, Theory of convex structures , V ol. 50 of North-Holland Mathematical Library, North-Holland, 1993

    1993

Pith tools

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