Hyper-Catalan and Geode Recurrences and Three Conjectures of Wildberger

Dean Rubine

Hyper-Catalan and Geode Recurrences and Three Conjectures of Wildberger

T0 review · 0 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read This paper proves three conjectured closed forms for the coefficients of the 'Geode', a factor in the hyper-Catalan generating series, by recurrence and induction.

desk verdict Solid, honest proof of Wildberger's Geode closed forms via a clean induction; the new Geode recurrence is the real contribution, and the soft spots are minor and fixable. read the letter →

arxiv 2507.04552 v1 pith:4Z365G4E submitted 2025-07-06 math.CO

Dean Rubine This is my paper · ORCID

classification math.CO MSC 05A1505A1905A10

keywords hyper-CatalannumbersGeodepolygonsubdivisionsgeneratingfunctionsclosedformFuss-Catalancombinatorialidentitiesrecurrence

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 hyper-Catalan numbers count the ways to subdivide a roofed polygon into specified numbers of triangles, quadrilaterals, and larger polygons. Their generating series $S$ admits the factorization $S-1=(t_2+t_3+t_4+\cdots)G$, whose factor $G$ is called the Geode; its coefficients are natural numbers, but their combinatorial meaning is unknown. This paper proves three conjectured closed forms for those coefficients in the cases where the dissections use one shape or two consecutive shapes. The proof builds on the Lesser Geode Sum Theorem and the Geode Recurrence, then runs by induction using the closed form for hyper-Catalan numbers. A reader should care because these formulas pin down an otherwise mysterious object and give exact computations where previously only recurrences or conjectures were available.

What carries the argument

The Lesser Geode Sum Theorem is the load-bearing device: for every nonzero type $m$, the hyper-Catalan number $C_m$ equals the sum of Geode coefficients $G_k$ over the 'lessers' $k$ obtained by subtracting one from each nonzero component (in symbols, $C_m = \sum_{k\in L(m)} G_k$). Together with the Geode Recurrence, which writes each $G_m$ as $C_{m+\vec{j}} - \sum_{k\in L(m+\vec{j})\setminus\{m\}} G_k$ for a chosen index $j$, this reduces Geode coefficients to integer combinations of hyper-Catalans. The closed-form hyper-Catalan formula of Erdelyi-Etherington supplies the explicit values, and an algebraic identity shows the conjectured $H$ obeys the same summation law, completing the induction.

What would settle it

Numerically evaluate the Bi-Tri formula at, say, $m=2, n=3$: the formula gives $G[2,3] = \frac{16!}{13 \cdot 6 \cdot 10! \cdot 2! \cdot 3!}$, and the recurrence expansion gives $G[2,3] = C[3,3] - C[4,2] + C[5,1] - C[6,0]$. A mismatch between these integers would disprove Theorem 9 instantly; the same direct comparison can be run for any $(k,m,n)$ against the expansion of Theorem 11.

Watch

Extended reading notes

Core claim

The central discovery is a closed form for Geode coefficients of types with two consecutive shapes: if a type consists of $m$ copies of a $k$-gon and $n$ copies of a $(k+1)$-gon, then the Geode coefficient is $H(k,m,n) = \frac{(km+(k+1)(n+1))!}{(k(m+n+1)+1)(m+n+1)((k-1)m+k(n+1))!\, m!\, n!}$. The $k=2$ instance, giving $G[m,n] = \frac{(2m+3n+3)!}{(2m+2n+3)(m+n+1)(m+2n+2)!\, m!\, n!}$ for triangles and quadrilaterals, is the simplest 'Bi-Tri' case. The same framework also proves the single-shape case, in which Geode coefficients reduce to Fuss numbers. These are the three conjectures, and the paper establishes them by showing that the conjectured formulas satisfy the same Lesser Geode Sum identity as the actual coefficients, with induction on the number of $(k+1)$-gons.

Load-bearing premise

The paper's claim that the Geode Recurrence always terminates for the largest-index choice is justified only as 'reasonably clear,' so the general claim that every Geode coefficient expands into a finite integer combination of hyper-Catalans rests on an informal termination argument; the three closed-form theorems themselves are proven directly by induction and do not depend on this termination.

Editorial extensions

If this is right

The two-consecutive-shape closed form yields exact values for the Geode Bi-Tri and Tri-Quad arrays, allowing direct computation of sequences that were previously only reachable by recurrence.
The same induction pattern is expected to generalize to any two shapes, though the paper notes that non-consecutive shapes (e.g. triangles and pentagons) produce coefficients with large prime factors, suggesting no simple ratio-of-factorials formula there.
The Geode Recurrence itself is a new computational tool: it expresses each Geode coefficient as a finite integer combination of hyper-Catalans, enabling calculation for larger types where closed forms are unknown.
The unusual binomial coefficient identities (Theorems 12 and 13) follow from equating the two different expressions for Geode coefficients, and the paper suggests they hold for a wide range of parameter values, giving a new family of combinatorial identities.

Reading between the lines

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

If the termination of the Geode Recurrence can be proven by a well-founded ordering on type vectors, the expansion of every Geode coefficient as an integer combination of hyper-Catalans would be fully rigorous; the present paper only sketches why the largest-index rule stops.
The large primes observed in the two non-consecutive shape (Bi-Quad) coefficients suggest that the general two-shape Geode coefficient is not a simple ratio of factorials; a search for a hypergeometric closed form may fail, and an interpretation of these coefficients as counting some class of objects would be more productive.
The identity family of Theorem 13 for integer $t$ including $t\le 1$ hints at an underlying rational-function identity; testing it for non-integer or complex $t$ with falling powers could reveal a broader analytic statement.
Because the single- and two-consecutive-shape coefficients are now explicit, matching them against sequence databases may identify the Geode with a known combinatorial family, or conversely, the new sequences may point to structures not yet in the database.

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 paper proves what it claims: Wildberger's three Geode closed-form conjectures are established by a clean induction from the Lesser Geode Sum theorem. Theorem 10 is the general two-consecutive-shapes case, with Theorem 9 as the triangle/quadrilateral special case. I checked the induction step—it uses only the Lesser Geode Sum split and an algebraic identity that holds. The proofs are elementary, and the paper explicitly credits Gessel and Amdeberhan–Zeilberger for independent proofs. So the central results stand.

What's actually new here is the Geode recurrence (Theorem 8), which expresses each Geode coefficient in terms of hyper-Catalans, and the systematic use of the Lesser Geode Sum to derive closed forms. The paper is honest about what it doesn't know: no general Geode formula, and no combinatorial interpretation. That's fair.

Soft spots are relatively minor. The termination claim for the recurrence after equation (44) is hand-wavy ("reasonably clear"). It can be fixed with an explicit measure—for the largest-index selection, each step lowers the sum of all components except the maximum, so well-foundedness is easy to prove. The paper should supply that. Also Theorems 12 and 13 are binomial identities stated without proof, with a pointer to Riedel's Math StackExchange proofs. They're not used in the main closed-form theorems, so this is a completeness issue, not a hole in the logic. The Section 10 material from Gessel is reported secondhand but clearly labeled and not central.

The citation pattern looks fine: the dependency on the Wildberger–Rubine theorem (Theorem 1) is legitimate because that's a published result with an independent proof sketch, and the paper flags the concurrent literature. No circularity.

This is a paper for enumerative combinatorics readers who care about Fuss–Catalan-type formulas and polynomial series solutions. It should be refereed, not desk-rejected. The fix list is short: prove termination, and either prove or properly cite the two binomial identities. If the editor wants a quick decision, I'd accept conditionally.

Referee Report

0 major / 5 minor

Summary. The paper studies the generating series S of hyper-Catalan numbers and the Geode factor G defined by S-1=(t2+t3+t4+...)G. It derives a hyper-Catalan recurrence (Theorem 2) and a Geode recurrence (Theorem 8), then uses these, together with the Erdelyi-Etherington closed form (Theorem 3), to prove three conjectures of Wildberger: closed forms for Geode coefficients for a single shape (Theorem 7), for the bi-triangular case (Theorem 9), and for the general two-consecutive-shape case (Theorem 10). The paper also presents a binomial-coefficient identity family (Theorems 12 and 13) without proof, reports OEIS data, and recounts independent proofs by Gessel and by Amdeberhan-Zeilberger.

Significance. The induction proofs of Theorems 9 and 10 are explicit and checkable, and the algebraic identities (48) and (51) reduce to routine common-denominator verification. The Geode recurrence is a genuinely useful new tool, and the paper is admirably honest about the unproved termination assertion in Section 5, the absence of a general closed form, and the existence of independent proofs of the conjectures. If the local gaps described below are repaired, the paper provides a sound elementary derivation of Wildberger's closed forms and a framework for further study of the Geode array.

minor comments (5)

[Section 5, after Eq. (44)] The termination of the Geode recurrence for the largest-index selection X(m) is only asserted as "reasonably clear." Since this underpins the claim that every Geode coefficient can be expanded as a finite integer combination of hyper-Catalans, please add a proof of termination; a simple measure is that after each replacement m -> m + e_{X(m)} - e_i the sum of all components except the new unique largest index strictly decreases. This is a local fix and does not affect the closed-form proofs of Theorems 9 and 10.
[Theorems 12 and 13] These are stated as theorems but no proof is given; the "routine manipulation" and "again substituting" are not enough. Since they are not used in the central proofs, a short proof, a citation to Riedel's proofs, or an explicit statement that they are computer-verified observations would be appropriate.
[Notation, Sections 2 and 6] The notation C3, C4, C31, etc. is used before the convention for single-index types is fully fixed; a sentence clarifying that, for example, C4 denotes C[4] (all triangles) would help the reader.
[Eqs. (48) and (51)] The algebraic identities used in the induction steps are dense; please include one line showing the common-denominator manipulation (or note the verification is by direct simplification), which would improve verifiability.
[Section 10] The text says "The four Geode conjectures were also proven..." while the abstract and title say three conjectures; please clarify whether the additional conjecture is Gessel's Theorem 14.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Theorems 9 and 10 are proved by induction from the Lesser Geode Sum and the closed-form hyper-Catalan formula, neither of which assumes the conjectures.

full rationale

The central results Theorems 9 and 10 are proved by induction from the Lesser Geode Sum (Theorem 6) and the Erdelyi-Etherington closed form for hyper-Catalans (Theorem 3). The Lesser Geode Sum itself is derived in the paper from the Geode factorization (Theorem 5), which is quoted from Wildberger and Rubine with a proof sketch; this prior result is not equivalent to the conjectures being proved, so the self-citation is a dependency rather than a circular loading. The induction base uses Theorem 7, which is itself derived from Theorem 6 rather than assumed. The algebraic identities (48) and (51) are explicit finite checks that connect H to the hyper-Catalan closed form. The only admitted gap, the termination of the general Geode recurrence under the largest-index choice X(m) after Equation (44), is stated as 'reasonably clear' and is not used in the proofs of the three conjectures. Consequently, no derivation step reduces by construction to its own input, and no fitted quantity is relabeled as a prediction.

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

No numerical parameters are fitted; the main inputs are two imported theorems and a combinatorial decomposition. The index function X(m) is an algorithmic choice, not a fitted value, and the Geode itself is inherited from Wildberger et al., not introduced here.

assumptions (4)

standard math Multinomial theorem and formal manipulation of multivariate power series.
Used throughout Section 2 to expand powers of the generating series S.
domain assumption Every subdigon decomposes uniquely as null or as ∇_k(s1,...,sk), giving S = 1 + t2 S^2 + t3 S^3 + ... (Equation (10)).
This combinatorial recursion is the basis of the geometric polynomial formula and hence of all later derivations.
domain assumption The generating series S is the zero of g(α)=1-α+t2α^2+t3α^3+... (Theorem 1, Wildberger-Rubine).
Imported theorem from the authors' prior work; the paper sketches but does not fully reprove it.
domain assumption Closed form C_m = (2m2+3m3+4m4+...)! / ((1+m2+2m3+3m4+...)! m2! m3! ...) (Theorem 3, Erdelyi-Etherington).
Used in the proof of the closed-form Geode theorems; cited to the 1940 literature.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Hyper-Catalan and Geode Recurrences and Three Conjectures of Wildberger." pith.science (2026). https://pith.science/paper/4Z365G4E

@misc{pith2026250704552,
  author       = {Pith},
  title        = {Pith review of: Hyper-Catalan and Geode Recurrences and Three Conjectures of Wildberger},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4Z365G4E}},
  note         = {Machine review of arXiv:2507.04552}
}

abstract

The hyper-Catalan number $C[m_2,m_3,m_4,\ldots]$ counts the number of subdivisions of a roofed polygon into $m_2$ triangles, $m_3$ quadrilaterals, $m_4$ pentagons, etc. Its closed form has been known since Erd\'elyi and Etherington, 1940. In 2025, Wildberger and Rubine showed its generating sum $\mathbf{S}[t_2,t_3,t_4,\ldots]$ is a zero of the general geometric univariate polynomial. We use that to derive a recurrence for hyper-Catalans, which expresses each in terms of other hyper-Catalans with smaller indices, generalizing the well-known Catalan convolution sum. Wildberger notes the factorization $\mathbf{S}-1=(t_2 + t_3 + t_4 + \ldots)\mathbf{G}$, where the factor $\mathbf{G}$ is called the Geode. We derive a recurrence that let us express the Geode coefficients in terms of other hyper-Catalan and Geode coefficients, and ultimately in terms of hyper-Catalans alone. We use it to prove three conjectures of Wildberger, all closed forms for special cases of elements of $\mathbf{G}$. While the recurrence allows us to expand each Geode coefficient as an integer combination of hyper-Catalans, enabling calculation, a closed-form for the general Geode coefficient remains unknown, as does what it counts.

Figures

Figures reproduced from arXiv: 2507.04552 by the authors.

**Figure 1.** The algebra of subdigons: ∇4 creates a subdigon with a central pentagon. pentagons, etc. is of type m = [m2, m3, m4, . . .]. Trailing zeros don’t change the type, and m = [ ], the vector of all zeros, is the type of |, the null subdigon. Sm is the multiset of subdigons of type m, and its size |Sm| is the hyper-Catalan number Cm, the number of subdigons of type m. The roof of a subdigon is a distinguished side of the… view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 3 Pith papers

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

Ordered trees and the Geode
math.CO 2025-07 conditional novelty 5.0 of 10

The coefficients of the Geode power series count leaves that appear before any internal node in the post-order traversal of ordered trees.
Lattice paths and the Geode
math.CO 2025-07 accept novelty 5.0 of 10

The Geode G equals (1 - sum_{n>=1} t_n (1+S+...+S^{n-1}))^{-1} and counts nonnegative lattice paths with steps -1,0,1,2,... under a natural weight.
Exercises for A Hyper-Catalan Series Solution to Polynomial Equations, and the Geode
math.CO 2025-07 accept novelty 4.0 of 10

A solved-exercise companion that fills in derivations for hyper-Catalan polynomial solutions and introduces the Tutrank and Jumbo Geode counting arrays.

Reference graph

Works this paper leans on

10 extracted references · 8 canonical work pages · cited by 3 Pith papers

[1]

https://www.arxiv.org/pdf/2506.17862

Tewodros Amdeberhan and Doron Zeilberger, Proofs of Three Geode Conjectures, ArXiv.org (2025). https://www.arxiv.org/pdf/2506.17862

arXiv 2025
[2]

Nicolas Fuss, Solutio quaestionis, quot modis polygonum n laterum in polygona m laterum, per diagonales resolui queat [Solution to the Question: In how many ways can a polygon with n sides be resolved by its diagonals?], Nova Acta Acad. Sci. Imp. Pet. 9 (1795), 243-251

work page
[3]

Erd´ elyi and I

A. Erd´ elyi and I. M. H. Etherington, Some problems of non-associative combinations (2), Edinburgh Mathematical Notes 32 (1940), vii-xii

work page 1940
[4]

Electronic Theses and Dissertations

Jennifer French, Vector Partitions, East Tennessee State University, 2018. Electronic Theses and Dissertations. Paper 3392

work page 2018
[5]

and Knuth Graham Donald E

Ronald L. and Knuth Graham Donald E. and Patashnik, Concrete mathematics: a foundation for computer science, Addison-Wesley, Reading, 1989

work page 1989
[6]

Knuth, The Art of Computer Programming, Volume 1 (3rd Ed.): Fundamental Algorithms, Addison Wesley Longman Publishing Co., Inc., USA, 1997

Donald E. Knuth, The Art of Computer Programming, Volume 1 (3rd Ed.): Fundamental Algorithms, Addison Wesley Longman Publishing Co., Inc., USA, 1997

work page 1997
[7]

S. R. Mane, Multiparameter Fuss–Catalan numbers with application to algebraic equations, arXiv.org, 2024

work page 2024
[8]

OEIS Foundation Inc., The On-Line Encyclopedia of Integer Sequences, 2024

work page 2024

Show all 10 references

[9]

Raney, Functional composition patterns and power series reversion, Trans am math soc 94 (1960), 441-451

George N. Raney, Functional composition patterns and power series reversion, Trans am math soc 94 (1960), 441-451

1960
[10]

N. J. Wildberger and Dean Rubine, A Hyper-Catalan Series Solution to Polynomial Equa- tions, and the Geode, The American Mathematical Monthly 132 (2025), no. 5, 383–402

2025

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [1]

https://www.arxiv.org/pdf/2506.17862

Tewodros Amdeberhan and Doron Zeilberger, Proofs of Three Geode Conjectures, ArXiv.org (2025). https://www.arxiv.org/pdf/2506.17862

arXiv 2025

[2] [2]

Nicolas Fuss, Solutio quaestionis, quot modis polygonum n laterum in polygona m laterum, per diagonales resolui queat [Solution to the Question: In how many ways can a polygon with n sides be resolved by its diagonals?], Nova Acta Acad. Sci. Imp. Pet. 9 (1795), 243-251

work page

[3] [3]

Erd´ elyi and I

A. Erd´ elyi and I. M. H. Etherington, Some problems of non-associative combinations (2), Edinburgh Mathematical Notes 32 (1940), vii-xii

work page 1940

[4] [4]

Electronic Theses and Dissertations

Jennifer French, Vector Partitions, East Tennessee State University, 2018. Electronic Theses and Dissertations. Paper 3392

work page 2018

[5] [5]

and Knuth Graham Donald E

Ronald L. and Knuth Graham Donald E. and Patashnik, Concrete mathematics: a foundation for computer science, Addison-Wesley, Reading, 1989

work page 1989

[6] [6]

Knuth, The Art of Computer Programming, Volume 1 (3rd Ed.): Fundamental Algorithms, Addison Wesley Longman Publishing Co., Inc., USA, 1997

Donald E. Knuth, The Art of Computer Programming, Volume 1 (3rd Ed.): Fundamental Algorithms, Addison Wesley Longman Publishing Co., Inc., USA, 1997

work page 1997

[7] [7]

S. R. Mane, Multiparameter Fuss–Catalan numbers with application to algebraic equations, arXiv.org, 2024

work page 2024

[8] [8]

OEIS Foundation Inc., The On-Line Encyclopedia of Integer Sequences, 2024

work page 2024