Pith. sign in

REVIEW 2 major objections 3 minor 33 references

Proof of a conjecture by Starr and log-concavity for random commuting permutations

T0 review · 2 major / 3 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read The number of ordered commuting tuples of permutations with a typical number of joint orbits has the exact asymptotic shape conjectured earlier, with explicit constants, for every ℓ ≥ 2.

desk verdict Proves Starr's conjecture with a genuine saddle-point argument; the log-concavity corollary has a real remainder gap and Lemma 2.3 has corrupted exponents, but the core is sound. read the letter →

arxiv 2506.06894 v1 pith:KCK4TEA6 submitted 2025-06-07 math.CO hep-thmath.NTmath.PR

Abdelmalek Abdesselam This is my paper
classification math.COhep-thmath.NTmath.PR MSC 05A1605A2011M0611M41
keywords commutingpermutationsjointorbitsStarrconjecturelog-concavitysaddlepointanalysisMellintransformDirichletseriesStirlingnumbers
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 the conjecture, due to Shannon Starr, that the counts $A(\ell,n,k)$ of ordered commuting $\ell$-tuples of permutations with $k$ joint orbits have a precise asymptotic form in the typical regime where $k$ grows like $k_n \sim s n^{(\ell-1)/\ell}$. The main theorems give $A(\ell,n,k) \sim \frac{\sqrt{\ell-1}}{2\pi}\,(n-1)!\,\exp(S_\ell(n,k)+k\,E_\ell(k/n))$ with fully explicit constants; the correction $E_\ell$ is a polynomial for $\ell\geq 4$ and involves logarithmic terms for $\ell=2,3$, exactly the structure the conjecture anticipated. A corollary is strict log-concavity in $k$ of these counts for large $n$ in that regime, the first such result for typical values of $k$. The proof also develops complete Mellin-transform asymptotics for the multiple sums $Z_{\alpha_1,\dots,\alpha_\ell}^{[\ell]}(t)$ that arise as the building blocks of the generating function, which may be of independent use.

What carries the argument

The proof packages the counts in the bivariate generating function $G_\ell(x,z)=\prod_{\delta_1,\dots,\delta_{\ell-1}\geq 1}(1-z^{\delta_1\cdots\delta_{\ell-1}})^{-x\,\delta_1^{\ell-2}\cdots\delta_{\ell-1}^0}$ and extracts $A(\ell,n,k)$ by two Cauchy integrals. The saddle point is controlled by the ratio $h_\ell(t)=(\ell-1)Z_{\ell-1}^{[\ell]}(t)/Z_\ell^{[\ell]}(t)$, whose monotonicity is guaranteed by a strict log-convexity lemma for the staircase $Z$ functions; the remaining prefactor asymptotics are obtained by inserting full Mellin-based $t\to 0^+$ expansions of the $Z$ functions and by reverting a formal power series through an explicit tree/Feynman-diagram formula.

What would settle it

Evaluate the difference $Z_m(t)^2-Z_{m-1}(t)Z_{m+1}(t)$ numerically for $\ell=2$ or $3$ at a few values of $t$ and negative $m$; a negative value would invalidate Lemma 2.3 and collapse the saddle point construction, since $h_\ell$ would not be an increasing bijection. Alternatively, compute $A(2,n,k)$ by exact enumeration for $n\approx 300$ and compare with Theorem 1.2: the predicted relative error is $O(n^{-1/2})$ and a systematic mismatch beyond it would refute the asymptotics.

Watch

Extended reading notes

Core claim

The central claim is that for each $\ell\geq 2$ the asymptotics of $A(\ell,n,k)$ in the window $k\sim s\,n^{(\ell-1)/\ell}$ are given, up to a factor $(n-1)!$ and an explicit constant, by $\exp(S_\ell(n,k)+k\,E_\ell(k/n))$, where $S_\ell$ is the leading entropy-like term and $E_\ell$ is a finite explicit correction whose shape changes with $\ell$: polynomial of degree $\ell-1$ for $\ell\geq 4$, with one $u\ln u$ term for $\ell=2$ and one $u^2\ln u$ term for $\ell=3$. This confirms the conjecture completely, including the logarithmic corrections that the original heuristic allowed only as possibilities, and shows no other logarithms appear for $\ell\geq 4$. The same asymptotic theorems imply the log-concavity inequality $A(\ell,n,k)^2 > A(\ell,n,k-1)A(\ell,n,k+1)$ for large $n$ in this regime.

Load-bearing premise

The entire construction hinges on the strict log-convexity of the functions $Z_m^{[\ell]}(t)$ in the staircase index $m$; if that inequality failed, the saddle ratio $h_\ell(t)$ would not have to be the increasing bijection $(0,\infty)\to(0,\ell-1)$ on which the choice of contours depends. The printed proof of this lemma displays corrupted exponents in the symmetrization step, so the inequality cannot be verified from the text as written.

Editorial extensions

If this is right

  • For all $\ell\geq 2$ and any $s\in(0,1)$, $A(\ell,n,k)^2 > A(\ell,n,k-1)A(\ell,n,k+1)$ holds for large $n$ whenever $k_n\sim s\,n^{(\ell-1)/\ell}$; this is the first log-concavity result in the typical-value regime.
  • The complete asymptotics of the $Z$ functions (Theorem 1.4) gives full $t\to 0^+$ expansions, including logarithmic terms from resonances, for all complex parameters; these multiple sums appear in generating functions of Dirichlet convolutions of power laws.
  • For $\ell\geq 4$ the correction $E_\ell$ contains no logarithmic terms, so the leading asymptotics are purely polynomial; the cases $\ell=2,3$ are exactly the ones carrying logarithms.
  • The bivariate saddle point analysis yields the needed prefactor unconditionally, without changing the topology of the integration torus.

Reading between the lines

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

  • If the log-concavity inequality holds beyond the typical window, the same saddle point construction might be adapted, since the saddle equations continue to make sense there; this would test the original log-concavity conjecture in full range.
  • The symmetrization argument behind Lemma 2.3 controls a single index direction $m$; the open question in Remark 2.1 about mixed differences of $Z$ along basis directions suggests a multivariable version of the inequality, which would bear on multivariate enumeration problems.
  • The $Z$-function asymptotics, being uniform in the parameters except at resonances, could be used to extract fine statistics of joint orbits beyond the central limit theorem proved by the companion paper, such as local limit theorems, by plugging oscillatory weights into the same saddle point.
  • The explicit constants, built from special values of $\zeta$ and $\zeta'$ at integers, should match those obtainable by a probabilistic re-derivation under the associated Ewens-type measure; a direct check of e.g. $I_{1,\log}=-3/\pi^2$ would tie the two approaches.
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

2 major / 3 minor

Summary. The manuscript proves Starr's conjecture on the asymptotic number A(ℓ,n,k) of ordered commuting ℓ-tuples of permutations with k joint orbits, in the typical regime k_n ∼ s n^{(ℓ−1)/ℓ}. The proof combines the Bryan–Fulman product formula with bivariate saddle point analysis, Mellin-transform asymptotics for the multiple sums Z_α^{[ℓ]}(t), and formal power series reversion for the prefactor. Theorems 1.1–1.3 give explicit asymptotic formulas for ℓ ≥ 4, ℓ = 2, and ℓ = 3, respectively, with the stated constants. As a corollary, the paper claims strict log-concavity of A(ℓ,n,k) in k for large n in the same typical regime. A separate theorem (Theorem 1.4) provides full asymptotic expansions for the Z functions.

Significance. If the main theorems are correct, they settle a conjecture of Starr and provide the first rigorous asymptotics for commuting tuples of permutations in the typical regime, generalizing classical results for Stirling numbers. The explicit nature of the constants and the absence of fitted parameters are strengths, as is the detailed Mellin analysis of the multivariate Z functions, which is likely to be useful independently. However, the proof of the log-concavity corollary has a load-bearing gap, and the crucial log-convexity lemma for the Z functions is not verifiable as written because of corrupted exponents. These issues are local and appear repairable, but they affect two of the paper's central claims.

major comments (2)
  1. [Section 6, proof of Corollary 1.1] The proof of Corollary 1.1 applies Theorems 1.1–1.3 to k_n, k_n−1 and k_n+1 and then writes Υ_n = ℓ Υ_{n,main} + Υ_{n,err}, retaining only the explicit S_ℓ and kE_ℓ terms. The theorems, however, only give log A = log C + log((n−1)!) + S_ℓ + kE_ℓ + o(1) for each of the three sequences; the three remainders contribute only o(1) to Υ_n, while the claimed main term is ∼ s^{-1} n^{-(ℓ−1)/ℓ}, which tends to 0. Since o(1) is not o(1/k_n), the displayed positivity argument does not control the remainder. Even the stronger log-level expansion log A = log C + log((n−1)!) + S_ℓ + kE_ℓ + O(k(k/n)^ℓ) + o(1) visible in Section 4 has remainder O(n^{-1/ℓ}), which for ℓ > 2 is much larger than 1/k_n and for ℓ = 2 is the same order; no second finite-difference bound for this remainder is supplied. The corollary therefore needs a strengthened, locally uniform version of the main theorems with an error term whose second finite difference is o(1/k_n), or an independent argument.
  2. [Section 2, Lemma 2.3] The displayed computation of Z_{m+1}Z_{m−1} − Z_m^2 does not match the definition (3). Since Z_m^{[ℓ]} = Z_{m,m−1,...,m−ℓ+1}, the product Z_{m+1}Z_{m−1} should have δ-exponents m+1, m, ..., m−ℓ+2 and η-exponents m−1, m−2, ..., m−ℓ, whereas the first line of the proof shows δ^m_1 δ^{m−1}_2 · · · δ^1_ℓ and η^{m−2}_1 · · · η^{-1}_ℓ; the subsequent factorization and symmetrization cannot be followed from the text. This lemma is load-bearing because it is used immediately afterward to prove h′_ℓ(t) > 0 and hence the bijectivity of h_ℓ that defines the saddle point. The proof should be rewritten with the correct exponents, or replaced by a Cauchy–Schwarz argument, which gives the stated inequality.
minor comments (3)
  1. [Throughout] There are several typographical errors: 'indepentent' in the abstract, 'oulined' in Section 1.3, 'Propostion' in Section 2, and 'transfom' in the title of reference [33].
  2. [Section 1.1] The phrase 'random commuting permutations' in the title and abstract is informal, since the paper is purely deterministic; this is not a mathematical issue but may mislead readers.
  3. [Section 6] In the finite-difference computation, the intermediate expansion for (1−1/k)ln(1−1/k)+(1+1/k)ln(1+1/k) is correct in its conclusion but would be clearer if the cancellation of the 1/k terms were shown explicitly.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: the asymptotics are derived from the Bryan–Fulman generating function, Mellin/saddle-point analysis, and standard reversion; self-citations are supplementary rather than load-bearing.

full rationale

The paper's central derivation is self-contained in the relevant sense. The starting point is the independent Bryan–Fulman infinite product (Eq. 4), from which the saddle-point equations, the prefactor M_ℓ(t_n,ρ_n), and the integral I_n are all derived in the text. The Z-function asymptotics used for the prefactor are proven in Theorem 1.4 and Proposition 2.1 via Mellin inversion and residue computation, not assumed. The constants K_ℓ, H_j, I_j, and J_j are explicit zeta-values and reversion coefficients, not fitted parameters, and the final E_ℓ(u) terms come from expanding h_ℓ(t) and composing with its inverse; the S_ℓ(n,k) expression is obtained by direct algebra from the saddle-point definitions (Eqs. 39, 7, 8, 26), not imported from the conjecture. The only repeated self-citations are to [6] for elementary Z-function properties and minor auxiliary estimates, and to [2] for the general Feynman-diagram reversion formula; these are independent published results, and the reversion theorem is a standard general tool whose assumptions do not include the target asymptotics. The log-concavity corollary is presented as a consequence of Theorems 1.1–1.3, not as a premise. I therefore find no circular step. A separate, non-circular correctness concern exists in the proof of Corollary 1.1: the application of Theorems 1.1–1.3 gives only o(1) remainders in the logarithms, and the finite-difference argument does not explicitly show these remainders are o(1/k_n), so the displayed positivity conclusion is not fully justified as written. That is a gap in the final inference, not a reduction of the theorem to its inputs, and it does not affect the circularity score.

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

No parameters are fitted to the data; the constants H, I, J are explicit expressions in special values of ζ and its derivatives. The only nonstandard inputs are the Bryan-Fulman formula (an existing theorem) and the author's reversion theorem; both are independent of the target result.

assumptions (4)
  • domain assumption The Bryan-Fulman product formula (4), G_ℓ(x,z)=∏_{δ_1,...,δ_{ℓ-1}≥1}(1-z^{δ_1...δ_{ℓ-1}})^{-x δ_1^{ℓ-2}...δ_{ℓ-2}}, extends to an analytic identity for complex x and |z|<1.
    This identity (quoted from [12] and [6]) is the bridge between A(ℓ,n,k) and the Z functions; the entire contour-integral representation in Section 1.3 rests on it.
  • standard math Standard Mellin transform and residue calculus, including the analytic continuation of ζ(s) and Γ(s), the pole structure, and uniform polynomial growth of ζ in vertical strips.
    This is the basis for Theorem 1.4's full asymptotic expansion of Z functions; cited from [7,8,14,29,31].
  • standard math The Lagrange-Good reversion theorem (from the author's earlier paper [2]) is used to compute the compositional inverse of h_ℓ and derive the D constants.
    A published theorem, but it is the author's own; the application to asymptotic power series reversion is standard and not circular in the sense of assuming the target result.
  • standard math The asymptotic expansion calculus for formal power series in C[[t]] (sums, products, composition, inversion) as in de Bruijn [11, Section 1.6].
    Used in Section 4 to convert the expansions of U, V, h_ℓ into the constants H, I, J.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Proof of a conjecture by Starr and log-concavity for random commuting permutations." pith.science (2026). https://pith.science/paper/KCK4TEA6

@misc{pith2026250606894,
  author       = {Pith},
  title        = {Pith review of: Proof of a conjecture by Starr and log-concavity for random commuting permutations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KCK4TEA6}},
  note         = {Machine review of arXiv:2506.06894}
}
read the original abstract

We prove a conjecture by Shannon Starr regarding the asymptotics for the number of tuples of commuting permutations with given number of joint orbits. These numbers generalize unsigned Stirling numbers of the first kind which count how many single permutations have a given number of cycles. In the case of pairs of permutations, these numbers are related to D'Arcais polynomials and the Nekrasov-Okounkov formula. As a consequence of the above asymptotics, we confirm a log-concavity conjecture in the regime of typical values for the number of joint orbits. As a result of possible indepentent interest in applied mathematics and mathematical physics, we also provide detailed asymptotics, using Mellin transform techniques, for certain multiple series or multivariate Ramanujan sums which are related to ordinary generating functions of Dirichlet convolutions of power laws. Besides these multiple sums asymptotics, our proofs use bivariate saddle point analysis related to the Meinardus theorem in the delicate case of multiple poles for the associated Dirichlet series.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

33 extracted references · 33 canonical work pages

  1. [1]

    Abdesselam, A physicist’s proof of the Lagrange-Good multivariable inversion formula

    A. Abdesselam, A physicist’s proof of the Lagrange-Good multivariable inversion formula. J. Phys. A 36 (2003), no. 36, 9471–9477

    work page 2003

  2. [2]

    Abdesselam, Feynman diagrams in algebraic combinatorics

    A. Abdesselam, Feynman diagrams in algebraic combinatorics. S´ em. Lothar. Combin. 49 (2002/04), Art. B49c, 45 pp. (electronic)

    work page 2002

  3. [3]

    Non-Abelian correlation inequalities and stable determinantal polynomials

    A. Abdesselam, Non-Abelian correlation inequalities and stable determinantal polynomials. Preprint arXiv:2207.07603[math-ph], 2022

    work page Pith review arXiv 2022

  4. [4]

    Abdesselam, Log-concavity with respect to the number of orbits for infinite tuples of commuting permutations

    A. Abdesselam, Log-concavity with respect to the number of orbits for infinite tuples of commuting permutations. Ann. Comb. 29 (2025), no. 2, 563–573. 29

    work page 2025

  5. [5]

    Abdesselam, P

    A. Abdesselam, P. Brunialti, T. Doan, and P. Velie, A bijection for tuples of commuting permutations and a log-concavity conjecture. Res. Number Theory 10 (2024), no. 2, Paper No. 45, 10 pp

    work page 2024

  6. [6]

    Abdesselam, and S

    A. Abdesselam, and S. Starr, A central limit theorem for a generalization of the Ewens measure to random tuples of commuting permutations. Preprint arXiv:2505.11469[math.PR], 2025

    work page arXiv 2025

  7. [7]

    G. E. Andrews, R. Askey, and R. Roy, Special functions. Encyclopedia of Mathematics and its Appli- cations 71. Cambridge University Press, Cambridge, 1999

    work page 1999

  8. [8]

    T. M. Apostol, Introduction to Analytic Number Theory . Undergrad. Texts Math., Springer-Verlag, New York-Heidelberg, 1976

    work page 1976

Show all 33 references
  1. [9]

    B. C. Berndt, and R. J. Evans, Extensions of asymptotic expansions from Chapter 15 of Ramanujan’s second notebook. J. Reine Angew. Math. 361 (1985), 118–134

    1985

  2. [10]

    Bringmann, J

    K. Bringmann, J. Franke, and B. Heim, Asymptotics of commuting ℓ-tuples in symmetric groups and log-concavity. Res. Number Theory 10 (2024), no. 4, Paper No. 83, 19 pp

    2024

  3. [11]

    N. G. de Bruijn, Asymptotic Methods in Analysis, Second edition. Bibl. Math., Vol. IV, North-Holland Publishing Co., Amsterdam; P. Noordhoff Ltd., Groningen, 1961

    1961

  4. [12]

    Bryan, and J

    J. Bryan, and J. Fulman, Orbifold Euler characteristics and the number of commuting m-tuples in the symmetric groups. Ann. Comb. 2 (1998), no. 1, 1–6

    1998

  5. [13]

    Debruyne, and G

    G. Debruyne, and G. Tenenbaum, The saddle-point method for general partition functions. Indag. Math. (N.S.) 31 (2020), no. 4, 728–738

    2020

  6. [14]

    H. M. Edwards, Riemann ’s Zeta Function. Reprint of the 1974 original (Academic Press, New York), Dover Publications, Inc., Mineola, NY, 2001

    1974

  7. [15]

    Flajolet, X

    P. Flajolet, X. Gourdon, and P. Dumas, Mellin transforms and asymptotics: harmonic sums. Special volume on mathematical analysis of algorithms Theoret. Comput. Sci. 144 (1995), no. 1–2, 3–58

    1995

  8. [16]

    Heim, and M

    B. Heim, and M. Neuhauser, Horizontal and vertical log-concavity. Res. Number Theory 7 (2021), no. 1, Paper No. 18, 12 pp

    2021

  9. [17]

    Hwang, Limit theorems for the number of summands in integer partitions

    H.-K. Hwang, Limit theorems for the number of summands in integer partitions. J. Combin. Theory Ser. A 96 (2001), no. 1, 89–126

    2001

  10. [18]

    Kowalski, Un Cours de Th´ eorie Analytique des Nombres

    E. Kowalski, Un Cours de Th´ eorie Analytique des Nombres. Cours Sp´ ec.,13, Soci´ et´ e Math´ ematique de France, Paris, 2004

    2004

  11. [19]

    J. C. Lagarias, Euler’s constant: Euler’s work and modern developments. Bull. American Math. Soc. (N.S.) 50 (2013), no. 4, 527–628

    2013

  12. [20]

    G. F. Lipnik, M. G. Madritsch, and R. F. Tichy, A central limit theorem for integer partitions into small powers. Monatsh. Math. 203 (2024), no. 1, 149–173

    2024

  13. [21]

    G. F. Lipnik, M. G. Madritsch, and R. F. Tichy, A local limit theorem for integer partitions into small powers. Preprint arXiv:2311.09203 [math.CO], 2023

    2023 arXiv

  14. [22]

    Madritsch, and S

    M. Madritsch, and S. Wagner, A central limit theorem for integer partitions. Monatsh. Math. 161 (2010), no. 1, 85–114

    2010

  15. [23]

    Moser, M

    L. Moser, M. Wyman, Asymptotic development of the Stirling numbers of the first kind. J. London Math. Soc. 33 (1958), 133–146

    1958

  16. [24]

    B. W. Ninham, B. D. Hughes, N. E. Frankel, and M. L. Glasser, M¨ obius, Mellin, and mathematical physics. Physica A 186 (1992), no. 3-4, 441–481

    1992

  17. [25]

    Pemantle, and M

    R. Pemantle, and M. C. Wilson, Analytic Combinatorics in Several Variables . Cambridge Stud. Adv. Math. 140, Cambridge University Press, Cambridge, 2013

    2013

  18. [26]

    Riemann, ¨Uber die Anzahl der Primzahlen unter einer gegebenen Gr¨ osse

    B. Riemann, ¨Uber die Anzahl der Primzahlen unter einer gegebenen Gr¨ osse. Monatsberichte der Berliner Akademie (1859) 671–680

  19. [27]

    generalized abundancy index

    S. Starr, Some observations about the “generalized abundancy index”. Preprint arXiv:2505.07051[math.CO], 2025

    2025 arXiv

  20. [28]

    Tao, Compactness and Contradiction

    T. Tao, Compactness and Contradiction . American Mathematical Society, Providence, RI, 2013

    2013

  21. [29]

    Tenenbaum, Introduction to Analytic and Probabilistic Number Theory

    G. Tenenbaum, Introduction to Analytic and Probabilistic Number Theory . Third edition. Translated from the 2008 French edition by Patrick D. F. Ion. Grad. Stud. Math. 163 American Math. Soc., Providence, RI, 2015

    2008

  22. [30]

    Tenenbaum, J

    G. Tenenbaum, J. Wu, and Y.-L. Li, Power partitions and saddle-point method. J. Number Theory 204 (2019), 435–445. 30

    2019

  23. [31]

    E. C. Titchmarsh, The Theory of The Riemann Zeta-Function . Second edition. Edited and with a preface by D. R. Heath-Brown. The Clarendon Press, Oxford University Press, New York, 1986

    1986

  24. [32]

    Wigert, Sur la s´ erie de Lambert et son application ` a la th´ eorie des nombres

    S. Wigert, Sur la s´ erie de Lambert et son application ` a la th´ eorie des nombres. Acta Math.41 (1916), no. 1, 197–218

    1916

  25. [33]

    Zagier, The Mellin transfom and related analytic techniques

    D. Zagier, The Mellin transfom and related analytic techniques. Appendix to Ch. 6 of the book by E. Zeidler Quantum Field Theory I: Basics in Mathematics and Physics. A Bridge Between Mathemati- cians and Physicists , pp. 305-323. Springer-Verlag, Berlin-Heidelberg-New York, 2...

    2006

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.