Asymptotically optimal lower bounds on weak saturation numbers for hypergraphs

Nikolai Terekhov

Asymptotically optimal lower bounds on weak saturation numbers for hypergraphs

T0 review · 1 major / 1 minor · reviewed 2026-07-13 · grok-4.5

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Weak saturation numbers for hypergraphs admit asymptotically optimal lower bounds from minimum degree, proved via polymatroids.

desk verdict Abstract-only claim of asymptotically optimal hypergraph weak-saturation lower bounds via polymatroids; real if true, but uncheckable without proofs. read the letter →

arxiv 2604.07104 v2 pith:UK45EBI6 submitted 2026-04-08 math.CO

Nikolai Terekhov This is my paper

classification math.CO MSC 05C6505C3505B35

keywords weaksaturationnumberr-uniformhypergraphspolymatroidsminimumdegreeasymptoticextremaldensitybootstrappercolation

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

The paper takes a classical extremal quantity for graphs—the weak saturation number wsat(n,H), the fewest edges an n-vertex host can have while still allowing every missing edge to be restored one-by-one so that each restoration creates a new copy of H—and extends the best known asymptotic lower bounds on that quantity from ordinary graphs to r-uniform hypergraphs. The bounds are expressed solely in terms of the minimum vertex degree of H; the author shows that the same functional form continues to hold for every uniformity r and that the resulting asymptotic coefficient is tight. The technical engine is a new lower-bound technique that replaces the usual linear-algebraic rank argument by a polymatroid rank function; the polymatroid framework recovers the integer-coefficient bounds of the linear-algebra method while also producing non-integer coefficients that match the true asymptotic density for a wider class of hypergraphs. A sympathetic reader therefore obtains a clean, degree-driven formula that is both general and asymptotically sharp for every fixed r-uniform H.

What carries the argument

A polymatroid rank function that lower-bounds the edge set of any weakly H-saturated r-uniform hypergraph; the construction generalizes the classical linear-algebraic method and yields asymptotic coefficients that need not be integers.

What would settle it

Exhibit a single fixed r-uniform H for which the asymptotic density of wsat(n,H)/n^{r-1} is strictly smaller than the coefficient predicted by the minimum-degree formula, or show that no polymatroid rank function can produce that coefficient.

Watch

Extended reading notes

Core claim

For every r-uniform hypergraph H the weak saturation number wsat(n,H) is asymptotically bounded from below by a coefficient that depends only on the minimum degree of H; the same coefficient is asymptotically tight, and the proof proceeds by constructing a suitable polymatroid whose rank lower-bounds the number of edges in any weakly H-saturated host.

Load-bearing premise

That a polymatroid rank function can always be defined so that it correctly lower-bounds every weakly H-saturated host and that the resulting asymptotic coefficient is tight for arbitrary H.

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

The one thing you need to know: this abstract claims a clean generalization of the known asymptotically optimal lower bounds on wsat(n,H) from graphs to r-uniform hypergraphs, plus a polymatroid method that can produce non-integer asymptotic coefficients, and it asserts those bounds are tight. That would be a solid, usable advance inside saturation theory if the proofs hold.

What looks new is exactly that package. The graph case was already settled; the contribution is the hypergraph extension and the switch from linear algebra to polymatroids so the coefficient need not be integer. The abstract is clear about the target quantity (minimum number of edges so that the missing ones can be ordered to each create a copy of H) and about the dependence on the minimum degree of H. No free parameters or circular fitting show up in the claim itself.

The soft spot is total and obvious: we have only the abstract. There is no definition of the rank function, no verification that it lower-bounds the edge count of every weakly H-saturated r-graph, and no matching upper-bound construction that would make the asymptotics optimal for general H. The stress-test note is right that nothing further can be isolated; the claim is untestable rather than contradicted. Pure-math abstracts of this type are often correct, but soundness is simply not checkable from what we have.

This is for people who already work on bootstrap percolation, weak saturation, or extremal hypergraph problems and who care about general lower-bound techniques. It is not a paper for a broad reading group until the proofs appear. I would send it to a serious referee once the full text is available; the stated result is important enough inside the subfield to deserve that time. I would not cite it yet, and I would not bring the abstract alone to reading group. If the polymatroid argument works as advertised, it is worth engaging; until then it is a promissory note.

Referee Report

1 major / 1 minor

Summary. The manuscript claims to generalize known asymptotically optimal lower bounds on the weak saturation number wsat(n,H) from the graph case (r=2) to r-uniform hypergraphs, expressing the bounds in terms of the minimum vertex degree of H, and to prove that the generalized bounds are asymptotically optimal. The abstract states that the proofs rely on a new lower-bound method based on polymatroids, which extends a linear-algebraic technique and permits non-integer asymptotic coefficients.

Significance. If the claimed generalization and the polymatroid method are correct, the work would supply a uniform asymptotic lower-bound framework for weak saturation numbers of hypergraphs, together with a matching optimality statement. The ability to obtain non-integer leading coefficients would be a genuine methodological advance over purely linear-algebraic arguments and could be of lasting use in extremal hypergraph theory. Credit is due for explicitly targeting asymptotic optimality rather than merely existential lower bounds.

major comments (1)

[Abstract (full text unavailable)] Only the abstract is available for review. The central claims rest on the existence of a polymatroid rank function that lower-bounds the number of edges in any weakly H-saturated r-uniform hypergraph and yields a coefficient that is asymptotically tight for general H. Without the full text (definitions of the rank function, the comparison with upper-bound constructions, and the asymptotic analysis), these load-bearing steps cannot be verified. Consequently no soundness determination is possible.

minor comments (1)

[Abstract] The abstract is clear and self-contained as a statement of intent, but a published version should include at least a sketch of the polymatroid construction and a precise statement of the main theorem (including the exact asymptotic coefficient) so that the claims can be checked.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity detectable from the abstract; claimed polymatroid lower bounds and asymptotic optimality are presented as independent generalizations, not as self-fitted or self-definitional results.

full rationale

Only the abstract is available. It defines wsat(n,H) in the standard way, states that known asymptotically optimal lower bounds for graphs (r=2) in terms of the minimum vertex degree of H are being generalized to r-uniform hypergraphs, and asserts that a new polymatroid-based lower-bound method yields those bounds and establishes their asymptotic optimality (including non-integer coefficients). No equations, fitted parameters, uniqueness theorems, or load-bearing self-citations appear in the abstract. There is therefore no exhibited reduction of a claimed prediction or first-principles result to its own inputs by construction. Residual dependence on prior definitions of weak saturation and on the known graph-case results is ordinary scientific context, not circularity under the stated criteria. With no full text, no circular step can be quoted or verified; the honest finding is score 0 and empty steps.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

Abstract-only review: no free parameters, invented particles, or ad-hoc numerical fits are visible. The work rests on standard definitions of weak saturation and on the existence of a suitable polymatroid rank function that encodes the edge-addition process; those are domain assumptions of extremal combinatorics and matroid theory, not new entities invented for the paper.

assumptions (3)

domain assumption Definition of weak saturation number wsat(n,H) for r-uniform hypergraphs (missing edges can be ordered so each creates a copy of H).
Standard combinatorial definition assumed throughout; stated in the abstract’s opening paragraph.
ad hoc to paper Existence of a polymatroid whose rank function lower-bounds the number of edges in any weakly H-saturated hypergraph and yields the claimed asymptotic coefficient.
The abstract introduces this as the key method; the precise rank axioms and their verification are not supplied in the abstract and are load-bearing for the lower bound.
domain assumption Known asymptotically optimal lower bounds for the graph case (r=2) in terms of minimum vertex degree of H.
Cited as prior literature that the paper generalizes; treated as given background.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Asymptotically optimal lower bounds on weak saturation numbers for hypergraphs." pith.science (2026). https://pith.science/paper/UK45EBI6

@misc{pith2026260407104,
  author       = {Pith},
  title        = {Pith review of: Asymptotically optimal lower bounds on weak saturation numbers for hypergraphs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UK45EBI6}},
  note         = {Machine review of arXiv:2604.07104}
}

abstract

Given an $r$-uniform hypergraph $H$ and a positive integer $n$, the weak saturation number $\mathrm{wsat}(n,H)$ is the minimum number of edges in an $r$-uniform hypergraph $F$ on $n$ vertices such that the missing edges in $F$ can be added, one at a time, so that each added edge creates a copy of $H$. For the case of graphs ($r = 2$), asymptotically optimal general lower bounds for these numbers in terms of the minimum vertex degree of $H$ are known. In this work, we generalize these bounds to the case of hypergraphs and establish their asymptotic optimality. To prove this, we introduce a lower bound method based on polymatroids. This method generalizes a linear algebraic method but, unlike the original version, makes it possible to derive lower bounds with non-integer asymptotic coefficients.

Review history (2 revisions) →

Discussion (0). Sign in to comment.

Pith tools

Reviewed July 13, 2026 · model on record in the stance chip above.