A fractional Helly theorem for set systems with slowly growing homological shatter function

Marguerite Bin

REVIEW 4 major objections 5 minor 12 references

A fractional Helly theorem for set systems with slowly growing homological shatter function

T0 review · 4 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 For families of sets in R^d, a slowly growing homological shatter function is enough for the fractional Helly theorem, even when the Radon number is unbounded.

desk verdict A short, honest note: the graded parameter framework is a real contribution, but the main theorem as stated is not yet proven because the proof outsources its central step to an unstated external theorem and the bounding function has a fixable but real defect. read the letter →

arxiv 2411.18605 v1 pith:IJZ2GJNO submitted 2024-11-27 cs.CG

Marguerite Bin This is my paper · ORCID

classification cs.CG MSC 52A3552A01

keywords topologicalcombinatoricsfractionalHellytheoremRadonnumberhomologicalshatterfunctiongradedVCdimensionBettinumbers

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 a fractional Helly theorem for families of sets in $\mathbb{R}^d$ whose intersection homology is controlled by a slowly growing 'homological shatter function', a hypothesis that does not force the classical Radon number to be finite. The main result, Theorem 3.1, says that for every $b,d \ge 0$ and $\alpha \in (0,1)$ there are $\beta > 0$ and $n_0$ such that if $\phi_{\mathcal{F}}^{(\lceil d/2 \rceil)}$ is bounded above by the paper's constructed function $\Psi_{d,b}$, then any finite subfamily with a positive fraction of intersecting $(d+1)$-tuples contains a positive fraction of members with a common point. The novelty is that this conclusion is reached through new graded analogues of the Radon, Helly, and colorful Helly numbers, whose growth is shown to be controlled by the homological shatter function. The proof chains two conversion steps: a cited threshold theorem converts many intersecting $(d+1)$-tuples into many intersecting $m_0$-tuples, and a clique-number lemma converts those into a large intersecting subfamily. If the theorem is correct, it is a concrete step toward the conjecture that polynomial growth of the homological shatter function implies a fractional Helly theorem.

What carries the argument

The load-bearing objects are the graded parameters $r^{(t)}(\mathcal{F})$, $h^{(t)}(\mathcal{F})$, $ch^{(t)}(\mathcal{F})$ — the suprema of the ordinary Radon, Helly, and colorful Helly numbers over all subfamilies of size at most $t$ — together with the homological shatter function $\phi_{\mathcal{F}}^{(h)}(t)$, which records the largest reduced Betti number in dimensions $0,\dots,h$ among all intersections of at most $t$ sets of $\mathcal{F}$. The proof's mechanism is the chain of inequalities (3) and (4), which bound these graded parameters by functions of $\phi_{\mathcal{F}}^{(\lceil d/2 \rceil)}(t)$, and the specifically constructed function $\Psi_{d,b}$, built by inverting the Radon bound $R_d(b') = r(b'+1,d)$, which is chosen so that the bounds satisfy the quantitative hypotheses of the conversion lemmas. In effect, $\Psi_{d,b}$ is a threshold: any family whose low-degree intersection homology grows no faster than this function is guaranteed to have bounded Helly number and bounded graded colorful Helly number at the needed scale.

What would settle it

Exhibit a family $\mathcal{F}$ of sets in $\mathbb{R}^d$ with $\phi_{\mathcal{F}}^{(\lceil d/2 \rceil)}$ bounded by $\Psi_{d,b}$ in which a positive fraction of $(d+1)$-tuples intersect but no positive fraction of $m_0$-tuples intersect, where $m_0 = m(r(b,d))$; such a family would refute the cited threshold theorem on which the proof depends. A concrete construction would be the minimal example showing the bridge from $(d+1)$-wise to $m_0$-wise intersections fails under only the low-degree Betti bound.

Watch

Extended reading notes

Core claim

The central claim is Theorem 3.1. For every $b,d \ge 0$ and every $\alpha \in (0,1)$, there exist $\beta > 0$ and $n_0$ such that every family $\mathcal{F}$ of sets in $\mathbb{R}^d$ with $\phi_{\mathcal{F}}^{(\lceil d/2 \rceil)}$ bounded from above by $\Psi_{d,b}$ has the fractional Helly property: every finite subfamily $\mathcal{F}'$ with $|\mathcal{F}'| \ge n_0$ and at least $\alpha \binom{|\mathcal{F}'|}{d+1}$ intersecting $(d+1)$-tuples contains at least $\beta |\mathcal{F}'|$ members with a common point. The proof introduces the graded parameters $r^{(t)}$, $h^{(t)}$, $ch^{(t)}$ — the suprema of the ordinary Radon, Helly, and colorful Helly numbers over subfamilies of size at most $t$ — and shows, via the Radon bound $r(\phi_{\mathcal{F}}^{(\lceil d/2 \rceil)}(t), d)$, that these graded numbers cannot grow too fast. The function $\Psi_{d,b}$ is chosen precisely so that this growth satisfies inequalities (5) and (6), which put the family inside the hypotheses of the two conversion lemmas. The conclusion that all members of the final subfamily intersect uses the bounded Helly number $h(\mathcal{F}) \le r_0$. Because $\Psi_{d,b}$ is unbounded, the family may have infinite Radon number.

Load-bearing premise

The whole argument rests on a cited but unstated theorem that many intersecting $(d+1)$-tuples imply many intersecting $m_0$-tuples whenever $\phi_{\mathcal{F}}^{(\lceil d/2 \rceil)}(m_0)$ is bounded; if that theorem actually needs control of all Betti numbers up to dimension $d$, the proof of Theorem 3.1 does not go through.

Editorial extensions

If this is right

Any family of sets in $\mathbb{R}^d$ with $\phi_{\mathcal{F}}^{(\lceil d/2 \rceil)}$ bounded by $\Psi_{d,b}$ satisfies the fractional Helly theorem, even when its Radon number is infinite.
The constants in the theorem depend only on $b$, $d$, and $\alpha$, so the conclusion is uniform across all families with this shatter-function bound.
The graded-parameter method extends the fractional Helly theorem from families with finite homological complexity to families with slowly growing homological complexity.
The construction of $\Psi_{d,b}$ shows the current proof reaches only very slow growth — roughly the iterated logarithm — so the polynomial-growth conjecture remains strictly out of reach of this technique.

Reading between the lines

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

The proof depends on the unstated threshold theorem; re-proving that theorem under the low-degree Betti bound $\phi^{(\lceil d/2 \rceil)}$ alone would make Theorem 3.1 self-contained and likely extend to other ambient spaces.
The graded-parameter framework is not specific to $\mathbb{R}^d$: any space with a Radon-type bound $r(\cdot,\cdot)$ would inherit the same fractional Helly theorem from the same argument.
Section 4's examples have very simple nerves (disjoint unions of cliques), so they do not stress the new theorem; a natural test case is a family where $\phi^{(h)}$ is bounded for some $h < \lceil d/2 \rceil$ but not at $\lceil d/2 \rceil$, to see whether the dimension threshold in the theorem is sharp.
The gap between the very slow $\Psi_{d,b}$ and the conjectured polynomial threshold suggests that a genuinely different bridge, not a refinement of the present conversion lemmas, will be needed to settle the conjecture.

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.

Desk Editor's Note

read the letter

Colleague,

This paper does something genuinely useful: it introduces graded versions of the Radon, Helly, and colorful Helly numbers for set systems, and shows how growth of the homological shatter function controls these graded parameters. That framework is reusable and is the right kind of idea. The paper also includes a nice characterization result (Lemma 4.1: every nondecreasing function is the homological shatter function of some family) and an example with logarithmically growing graded Radon numbers that shows why the non-stationary case is worth separating from the bounded-homological-complexity setting. The writing is clear and the citations look right.

The main theorem (Theorem 3.1) is plausible, but I can't call it proven as written. Two problems, one small and one load-bearing.

Small: the function Ψ_{d,b} is not well-defined at t = r(b,d), where the first and second cases both apply and give b−1 and b. Worse, it can drop from b to S(t) immediately after m(r(b,d))·r(b,d), even though the homological shatter function it is meant to bound is nondecreasing in t. This is fixable by redefining the intervals (e.g., half-open) and by making Ψ nondecreasing, but the statement needs that repair before the hypothesis makes sense.

Load-bearing: the proof's only bridge from a positive fraction of intersecting (d+1)-tuples to a positive fraction of intersecting m0-tuples is [2, Theorem 1.2]. The paper never states that theorem or its hypotheses. The proof just says that φ^⌈d/2⌉_F(m0) ≤ b0 lets you apply it. But the theorem might require control of all Betti numbers up to dimension d, or a forbidden-homological-minor condition, or something else entirely. Without seeing the theorem, I can't verify that every family with φ^⌈d/2⌉ bounded by Ψ_{d,b} satisfies its hypotheses. This is the kind of gap that can be filled by quoting the theorem and checking the condition, but until then the central claim is unproved.

The examples in Section 4 are honest: the authors admit their construction does not need the full strength of the theorem and that only the ⌈d/2⌉-th shatter function is controlled. That is fine.

Who is this for? People working in topological combinatorics and fractional Helly theory. They will want the graded-parameter definitions regardless of the fate of Theorem 3.1. The paper deserves a serious referee, but the referee should require the author to state [2, Theorem 1.2] and verify its hypotheses, and to clean up Ψ. It is close to being a good paper, but not there yet.

Referee Report

4 major / 5 minor

Summary. This paper introduces graded analogues of the Radon, Helly, and colorful Helly numbers of a set system, and relates them to the homological shatter function. The main result (Theorem 3.1) claims that if the homological shatter function of order ceil(d/2) is bounded above by an explicitly defined slowly growing function Psi_{d,b}, then the family satisfies a fractional Helly theorem: a positive fraction of intersecting (d+1)-tuples forces a positive fraction of members with a common point. The proof combines inequalities of Patakova and of Holmsen-Lee with an external theorem [2, Theorem 1.2] that is used as a black box. The paper concludes with constructions of set systems with prescribed homological shatter functions and with logarithmically growing graded Radon numbers.

Significance. If the main theorem is correct, it extends fractional Helly theorems to families whose Radon number may be unbounded, giving a concrete step toward the Kalai-Meshulam conjecture. The graded-parameter framework is natural, Proposition 2.2 provides a clean general bound on graded Radon numbers, and Lemma 4.1 gives an explicit realization of arbitrary nondecreasing homological shatter functions. These are genuine contributions. However, the proof of the central theorem currently depends on an unstated external result and on an ill-defined bounding function, so the significance is conditional on repairing the gaps described below.

major comments (4)

[Section 3, proof of Theorem 3.1 (after fixing n0)] The only bridge from a positive fraction of intersecting (d+1)-tuples to a positive fraction of intersecting m0-tuples is the invocation of [2, Theorem 1.2], but that theorem is neither stated nor paraphrased. Consequently, the hypotheses that the family F must satisfy are never checked; for example, [2, Theorem 1.2] might require bounds on all Betti numbers up to dimension d, a forbidden homological minor, or a different relation between m0 and the shatter function. Since this transfer is the load-bearing step, the proof is incomplete as written. Please state [2, Theorem 1.2] and verify its hypotheses from the assumption phi^{ceil(d/2)}_F <= Psi_{d,b}.
[Section 3, definition of Psi_{d,b}] The piecewise definition of Psi_{d,b} has overlapping intervals at t = r(b,d), where the first branch assigns the value b-1 and the second assigns b, so Psi_{d,b} is not a well-defined function at that point. Moreover, unless one proves that m(r(b,d))r(b,d) >= r(b+1,d), the third branch can assign a value S(t) < b immediately after m(r(b,d))r(b,d); since phi is nondecreasing, no nondecreasing function can be bounded above by Psi_{d,b} in that regime, which would make the hypothesis of Theorem 3.1 vacuous for all families. Please repair the definition with disjoint intervals and either prove that Psi_{d,b} is nondecreasing or add this as an explicit condition.
[Section 3, proof of Theorem 3.1 (application of Lemma 2.3)] The conclusion of Lemma 2.3, as stated in the manuscript, is that every m0 sets of G intersect, not that every r0 sets of G intersect. The deduction that all elements of G intersect requires first passing from m0-wise intersection to r0-wise intersection, which needs m0 >= r0; no such inequality is established. Also, the argument that h(F) <= r0 turns r0-wise intersection into total intersection is only valid if |G| >= r0, so the choice of n0 should explicitly ensure that beta(alpha',r0,m0) n0 >= r0. Please either prove these inequalities or reformulate the application so that it matches the stated hypotheses and conclusion of Lemma 2.3.
[Section 3, Theorem 3.1 statement and proof] The proof uses the notation b0 and sets r0 = r(b0,d), while the theorem is stated with a fixed parameter b; this mismatch makes it impossible to see which value of b is being used in the final beta(b,d,alpha). Please unify the notation throughout the statement and proof.

minor comments (5)

[Abstract and Introduction] There are several typos, including 'growi ng', 'bounde d', 'speci c', and 'untrue'; a careful proofreading pass is needed.
[Section 3, proof of Theorem 3.1] The proof refers to 'Theorem 2.3' when the cited statement is 'Lemma 2.3'; please correct the cross-reference.
[Section 3, proof of Theorem 3.1] In the final display, beta(alpha',r0,m) should presumably be beta(alpha',r0,m0); please clarify the argument.
[Introduction and References] The text attributes the conjecture to 'Kalai and Meshulam in [7]', but reference [7] is listed as a single-author paper by Kalai; either correct the attribution or add the appropriate joint reference.
[Section 4, Lemma 4.1] The phrase 'for i < d-h, we can adapt the construction' is vague; please specify the adaptation or remove the clause.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Theorem 3.1 is a combination of independent external results, and the tailored bounding function Ψ is a legitimate sufficient-condition construction rather than a fitted prediction or self-citation.

full rationale

The derivation chain of Theorem 3.1 is a composition of independent external results: [11, Thm 2.1] bounds graded Radon numbers by r(φ^{(⌈d/2⌉)}(t),d), [6, Lemma 2.3] bounds colorful Helly numbers, [2, Thm 1.2] converts a positive fraction of intersecting (d+1)-tuples into intersecting m0-tuples under a homological shatter bound, and [5, Thm 1.2] converts clique bounds into a fractional subfamily. The function Ψ_{d,b} is explicitly defined so that any family bounded by it satisfies the two numeric inequalities (5) and (6) needed by the proof; this is a sufficient-condition construction, not a parameter fitted to the conclusion, and no prediction is renamed as an input. There are no self-citations by the author and no uniqueness import. The main weakness is that [2, Theorem 1.2] is cited without stating its hypotheses, so the proof as written does not verify them; this is an unproved-support/correctness risk (and the definition of Ψ has an overlap at t=r(b,d)), but neither makes the argument circular.

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

No physical entities are introduced. The graded Radon, Helly, and colorful Helly numbers are new mathematical definitions, and they are the content of the paper rather than unexplained postulates. The function Psi_{d,b} is a defined auxiliary bound, not an independent entity.

assumptions (4)

standard math Patakova's Radon bound [11, Theorem 2.1]: the Radon number of a family of sets in R^d is bounded by a function of its homological complexity.
Used to obtain the inequality h(t)(F) < r(t)(F) <= r(phi_{ceil(d/2)_F}(t), d) in Section 3, Inequality (3).
standard math Goaoc-Holmsen-Patakova [2, Theorem 1.2]: a positive fraction of intersecting (d+1)-tuples can be lifted to a positive fraction of intersecting m0-tuples under the bound phi(m0) <= b0.
Used to pass from dense (d+1)-intersections to dense m0-intersections in the proof of Theorem 3.1; its exact hypotheses are not quoted in the paper.
standard math Holmsen-Lee [6, Lemma 2.3] and Holmsen [5, Theorem 1.2]: colorful Helly bounds control clique numbers in hypergraphs.
Used for Inequality (2) and for Lemma 2.3, which is the main engine for the clique conclusion in Theorem 3.1.
standard math Alexander duality over Z2 in the construction of Lemma 4.1.
Used to compute the h-th Betti number of the constructed intersections in Section 4.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A fractional Helly theorem for set systems with slowly growing homological shatter function." pith.science (2026). https://pith.science/paper/IJZ2GJNO

@misc{pith2026241118605,
  author       = {Pith},
  title        = {Pith review of: A fractional Helly theorem for set systems with slowly growing homological shatter function},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IJZ2GJNO}},
  note         = {Machine review of arXiv:2411.18605}
}

abstract

We study parameters of the convexity spaces associated with families of sets in $\mathbb{R}^d$ where every intersection between $t$ sets of the family has its Betti numbers bounded from above by a function of $t$. Although the Radon number of such families may not be bounded, we show that these families satisfy a fractional Helly theorem. To achieve this, we introduce graded analogues of the Radon and Helly numbers. This generalizes previously known fractional Helly theorems.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

12 extracted references · 12 canonical work pages

[1]

J. P. Doignon. Convexity in crystallographical lattices. J. Geom. 3, 71–85 (1973)

work page 1973
[2]

Intersection patterns in spaces with a forbidden homological minor

X. Goaoc, A. Holmsen, and Z. Pat´ akov´ a. Intersection patterns in spaces with a forbidden homo- logical minor. arXiv e-prints (2024) arXiv:2103.09286

work page Pith review arXiv 2024
[3]

Goaoc, P

X. Goaoc, P. Pat´ ak, Z. Pat´ akov´ a, M. Tancer, and U. Wagner. Bounding Helly numbers via Betti numbers. In A journey through discrete mathematics , pages 407–447. Springer, Cham (2017)

work page 2017
[4]

E. Helly. ¨Uber systeme von abgeschlossenen mengen mit gemeinschaftlichen punkten. Monatsh. f. Mathematik und Physik 37, 281–302 (1930)

work page 1930
[5]

A. F. Holmsen. Large cliques in hypergraphs with forbidden subst ructures. In Combinatorica, volume 40, pages 527-537 (2020)

work page 2020
[6]

A. F. Holmsen and D. Lee. Radon numbers and the fractional Helly theorem. Isr. J. Math. 24, 433–447 (2021)

work page 2021
[7]

G. Kalai. Combinatorial expectations from commutative algebra. In I. Peeva and V. Welker, editors, Combinatorial Commutative Algebra , volume 1(3), pp 1729–1734. Oberwolfach Reports (2004)

work page 2004
[8]

Kay and E

D. Kay and E. Womble. Axiomatic convexity theory and relationship s between the Carath´ eodory, Helly, and Radon numbers. In Paciﬁc Journal Of Mathematics , volume 38, pp. 471-485 (1971)

work page 1971

Show all 12 references

[9]

F. Levi. On Helly’s theorem and the axioms of Convexity. In Journal of the Indian Mathematical Society, volume 15, pp. 65-76 (1951)

1951
[10]

Matouˇ sek

J. Matouˇ sek. Lectures on discrete geometry , volume 212. Springer Science & Business Media (2013)

2013
[11]

Pat´ akov´ a

Z. Pat´ akov´ a. Bounding Radon numbers via Betti numbers. I nternational Mathematics Research Notices (2024). https://doi.org/10.1093/imrn/rnae056

2024 doi
[12]

J. Radon. Mengen konvexer k¨ orper, die einen gemeinsamen pu nkt enthalten. Mathematische Annalen, volume 83, pp. 113-115 (1921). 7

1921

Pith tools

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

[1] [1]

J. P. Doignon. Convexity in crystallographical lattices. J. Geom. 3, 71–85 (1973)

work page 1973

[2] [2]

Intersection patterns in spaces with a forbidden homological minor

X. Goaoc, A. Holmsen, and Z. Pat´ akov´ a. Intersection patterns in spaces with a forbidden homo- logical minor. arXiv e-prints (2024) arXiv:2103.09286

work page Pith review arXiv 2024

[3] [3]

Goaoc, P

X. Goaoc, P. Pat´ ak, Z. Pat´ akov´ a, M. Tancer, and U. Wagner. Bounding Helly numbers via Betti numbers. In A journey through discrete mathematics , pages 407–447. Springer, Cham (2017)

work page 2017

[4] [4]

E. Helly. ¨Uber systeme von abgeschlossenen mengen mit gemeinschaftlichen punkten. Monatsh. f. Mathematik und Physik 37, 281–302 (1930)

work page 1930

[5] [5]

A. F. Holmsen. Large cliques in hypergraphs with forbidden subst ructures. In Combinatorica, volume 40, pages 527-537 (2020)

work page 2020

[6] [6]

A. F. Holmsen and D. Lee. Radon numbers and the fractional Helly theorem. Isr. J. Math. 24, 433–447 (2021)

work page 2021

[7] [7]

G. Kalai. Combinatorial expectations from commutative algebra. In I. Peeva and V. Welker, editors, Combinatorial Commutative Algebra , volume 1(3), pp 1729–1734. Oberwolfach Reports (2004)

work page 2004

[8] [8]

Kay and E

D. Kay and E. Womble. Axiomatic convexity theory and relationship s between the Carath´ eodory, Helly, and Radon numbers. In Paciﬁc Journal Of Mathematics , volume 38, pp. 471-485 (1971)

work page 1971