Pith. sign in

REVIEW 2 major objections 5 minor 5 references

The imaginary case of the nonabelian Cohen--Lenstra heuristics

T0 review · 2 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read For imaginary Γ-extensions of function fields, the average number of equivariant surjections from the maximal split-complete unramified Galois group to a finite admissible H is exactly $1/[H^{\Gamma_\infty}:H^\Gamma]$, and a random group…

desk verdict Solid new moment theorem and a plausible, explicitly conditional probability conjecture; the random-group foundation rests on an unproven generation assumption. read the letter →

arxiv 2507.21558 v1 pith:PRJPILZ3 submitted 2025-07-29 math.NT

Yuan Liu , Ken Willyard This is my paper · ORCID
classification math.NT MSC 11R2911R3211R5814H30
keywords nonabelianCohen-LenstraheuristicsimaginaryextensionsHurwitzstacksmomentcomputationsrandomgroupsGaloisfunctionfields
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 extends the nonabelian Cohen–Lenstra heuristics to the imaginary case, where the distinguished place ∞ is totally ramified instead of split. In the function field setting, it proves that the average number of Γ-equivariant surjections from $G^\#_\emptyset(K)$—the Galois group of the maximal unramified extension split completely at ∞—onto a fixed finite admissible Γ-group H, with a fixed ω-invariant pushforward, tends to $1/[H^{\Gamma_\infty}:H^\Gamma]$ as the radical discriminant grows. It also proves that these Galois groups admit a specific presentation, uses that presentation to build a random Γ-group whose moments match the computed limits, and conjectures that this random group governs the distribution over imaginary extensions of both $\mathbb{F}_q(t)$ and $\mathbb{Q}$. If true, the work gives a uniform prediction for the distribution of class-group-like objects in the imaginary case, completing the imaginary analogue of the earlier totally real results.

What carries the argument

The argument is carried by four coordinated pieces. First, the Hurwitz stack $\mathrm{Hur}^n_{G,G_\infty,c}$, a closed substack of the stack of tame Galois $G$-covers of $\mathbb{P}^1$ with $n$ branch points and a marked point over ∞; its reduced substack is smooth, and its $\mathbb{F}_q$-points correspond to the pairs $(K,\pi)$ counted in the moment. Second, the algebraic lifting invariant for Hurwitz components, which labels connected components and controls the Frobenius action inside the Lefschetz trace formula for algebraic stacks. Third, the presentation theorem: under a finite admissible generation hypothesis, $G^\#_\emptyset(K)_C$ is isomorphic to a free admissible pro-$C$ Γ-group on $n$ generators modulo the Γ-closed normal subgroup generated by $r_i^{-1}\gamma(r_i)$ for $i=1,\dots,n+1$, with $r_{n+1}$ fixed by $\Gamma_\infty$. Fourth, the random Γ-group $X_{\Gamma,\Gamma_\infty}$ built by drawing $x_1,\dots,x_n$ from the Haar measure on the free admissible group and $x_{n+1}$ from its $\Gamma_\infty$-invariants, whose moment formula yields the same reciprocal index.

What would settle it

To test the moment claims, compute the left-hand side of (1.1) for a small group Γ, a nontrivial cyclic $\Gamma_\infty$, and an admissible H with $[H^{\Gamma_\infty}:H^\Gamma]\neq 1$, over many large prime powers $q$ and increasing $n$; if the averages do not converge to the reciprocal index, Theorem 1.1 falls. To test the structural premise, search for an imaginary $(\Gamma,\Gamma_\infty)$-extension $K$ and an admissible $C$ for which $G^\#_\emptyset(K)_C$ is not finitely admissibly generated; such an example would invalidate the presentation theorem and the probability version of Conjecture 7.1.

Watch

Extended reading notes

Core claim

The central discovery is that the imaginary case of the nonabelian Cohen–Lenstra heuristics is controlled by the index $[H^{\Gamma_\infty}:H^\Gamma]$: the $H$-moment of the distribution of $G^\#_\emptyset(K)$ equals the reciprocal of this index. The paper proves this moment limit over function fields by counting $\mathbb{F}_q$-points on a closed substack of the Hurwitz stack that parametrizes covers ramified at the marked point over ∞, with the ω-invariant fixed. For the probability side, the paper proves that the pro-$C$ completion of $G^\#_\emptyset(K)$ is presented by a free admissible Γ-group on $n$ generators with $n+1$ relations $r_i^{-1}\gamma(r_i)$, where $r_{n+1}$ is fixed by $\Gamma_\infty$; this presentation is the basis of the random group $X_{\Gamma,\Gamma_\infty}$ whose moments are computed explicitly and equal $1/[H^{\Gamma_\infty}:H^\Gamma]$. Conjecture 7.1 asserts that this random group predicts the distribution over imaginary Γ-extensions of both $\mathbb{F}_q(t)$ and $\mathbb{Q}$.

Load-bearing premise

The presentation theorem assumes that the pro-$C$ completion of $G^\#_\emptyset(K)$ is finitely admissibly generated; the paper states that whether the underlying maximal split-complete unramified Galois group is finitely generated is an open question, so if that generation fails for some $K$, the random-group presentation would not describe the true Galois group.

Editorial extensions

If this is right

  • For function fields, Theorem 1.1 yields an explicit, testable average: over many prime powers $q$, the mean number of $H$-quotients of $G^\#_\emptyset(K)$ with fixed ω-invariant approaches $1/[H^{\Gamma_\infty}:H^\Gamma]$; this can be checked numerically for small Γ and H.
  • The random group $X_{\Gamma,\Gamma_\infty}$ supplies a candidate probability measure for the distribution over number fields; since moments determine measures in these settings, Conjecture 7.1 reduces the full distribution to the computed moment sequence.
  • The presentation theorem describes the structure of the maximal split-complete unramified extension's Galois group in terms of $\Gamma_\infty$-fixed generators and relations, giving a finiteness-type statement that did not previously exist for the imaginary case.
  • Specializing $\Gamma=\Gamma_\infty=\mathbb{Z}/2\mathbb{Z}$ recovers the classical Cohen–Lenstra moment $1$ for imaginary quadratic fields and agrees with the Boston–Bush–Hajir moment for odd $p$-class tower groups, so the conjecture is compatible with all previously known imaginary-case heuristics.

Reading between the lines

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

  • If the finite-generation assumption behind Theorem 1.2 holds for all imaginable Γ-extensions, the random group $X_{\Gamma,\Gamma_\infty}$ could be used to produce explicit numerical predictions for class group distributions of small imaginary number fields, which table computations could then test; the paper does not run such numerics.
  • The same ramified-at-∞ Hurwitz-stack counting may apply to other statistics of global fields with a prescribed totally ramified place, such as distributions with local conditions at multiple places, though the paper only treats the single-place case.
  • The paper leaves the roots-of-unity case (when the base field contains extra roots of unity) out of the probability version; a natural extension would be to add a pairing datum to the random group, in the spirit of earlier class-group moment work, to cover those fields.
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 / 5 minor

Summary. This paper develops the imaginary analog of the nonabelian Cohen--Lenstra heuristics. For a finite group Γ with a nontrivial cyclic subgroup Γ∞, it studies G#_∅(K), the pro-prime-to-Δ_Q completion of the Galois group of the maximal unramified extension of K split completely above ∞, as K ranges over imaginary (Γ,Γ∞)-extensions of F_q(t) or Q. Theorem 1.1 proves, by counting F_q-points on Hurwitz stacks and using lifting invariants, that the average number of Γ-equivariant surjections from G#_∅(K) to a finite admissible H with a fixed ω-invariant pushforward is 1/[H^{Γ∞}:H^Γ]; the fixed-q strengthening (1.2) is obtained by importing homological stability results of Landesman--Levy. Theorem 1.2 gives a presentation of pro-C completions of G#_∅(K) under a finite-admissible-generation hypothesis. From this presentation the authors construct a random Γ-group X_{Γ,Γ∞}, compute its probability measure and moments (Theorem 6.3), and conjecture that it governs the arithmetic distribution (Conjecture 7.1). They also verify compatibility with several existing conjectures and prove a weighted Gerth-type moment in §7.4.

Significance. If the main theorem is correct, this is a substantial advance: it gives an unconditional large-q moment theorem for imaginary extensions (1.1), a fixed-q version conditional on [LL25], and a concrete random-group model whose moments provably match the function-field computation with no fitted parameters. The compatibility checks with Cohen--Lenstra, Friedman--Washington, Cohen--Lenstra--Martinet, Boston--Bush--Hajir, and the Gerth-type weighted moment are valuable and give independent evidence that the proposed measure is the right one. However, the probability-level conjecture inherits the unproved finite-generation hypothesis of Theorem 1.2 for infinite sets C, so the full conjecture as stated is not yet supported by the structural theorem.

major comments (2)
  1. [§5.2 / Theorem 1.2 / Conjecture 7.1, Probability Version] The probability version of Conjecture 7.1 is stated for an arbitrary set C of finite Γ-groups, but the presentation theorem that justifies the random-group measure requires condition (a), namely that G#_∅(K)_C is finitely admissibly generated. The only families for which the paper proves (a) are finite C and p-class tower quotients (the paragraph after Theorem 1.2), and Section 5.2 itself records that finite generation of G∞_∅(K) is open. For an infinite C of mixed order the hypothesis is neither proved nor conjectured to hold with density 1; Remark 7.2(2) only concerns condition (b). The conjecture should be restricted to C satisfying (a), or (a) should be added as an explicit hypothesis of the conjecture, with a discussion of what would happen if (a) fails for a positive-density family of K.
  2. [Theorem 1.1, eq. (1.2)] The fixed-q moment statement is not proved from first principles in this paper but by importing homological-stability estimates from the unpublished preprint [LL25] (§4.2, especially equations (4.12)–(4.15)). Since (1.2) is presented as part of Theorem 1.1 and feeds the function-field moment conjecture, the paper should quote the exact statements needed from [LL25] and label (1.2) as conditional on that work; as it stands, the result cannot be independently checked from the present manuscript.
minor comments (5)
  1. [§4, first paragraph] “Hurtiwz stacks” should be “Hurwitz stacks”.
  2. [§1.1, paragraph after Theorem 1.1] “Landensman and Levy” should be “Landesman and Levy”.
  3. [Lemma 2.3 and surrounding text] The notation Z[|G|−1] is confusing; please write Z[|G|^{-1}] or Z[1/|G|].
  4. [Conjecture 7.1, Moment Version] The notation H2(H⋊Γ,Z)_{(|Γ|)'}[q−1] is used without definition; please state explicitly what [q−1] denotes and explain the factor in the formula.
  5. [§7.4, equation (7.1)] The notation “2 Cl(K)[2∞]” is not defined in the text; please clarify whether it denotes the quotient of Cl(K)[2∞] by its 2-divisible subgroup or another variant.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the Hurwitz-stack moment computation and the random-group moment calculation are independent derivations whose agreement is genuine evidence; the open finite-generation condition is a caveat, not a circular reduction.

full rationale

The paper's main proved result, Theorem 1.1, is a genuine point-count: Lemma 4.1 bijectively relates H-moments to F_q-points on the imaginary Hurwitz substack Hur^n_{G,G∞,c}, and Sections 3-4 estimate those point counts by Behrend's trace formula and lifting invariants, with no free parameters fitted to the advertised limit 1/[H^{Γ∞}:H^Γ]. The random-group moment, Theorem 6.3(3), is computed from Definition 6.1 and Lemma 5.3 directly; it is not obtained by calibrating the model to Theorem 1.1. The agreement between the two independent computations is meaningful corroboration. Theorem 1.2 is conditional on its hypothesis (a), and Section 5.2 explicitly says: "Note that it is an open question to determine whether G∞_∅(K) is finitely generated." Remark 7.2(2) conjectures only condition (b) for 100% of K, leaving (a) unproved for the infinite sets C used in Conjecture 7.1. This is a real foundational caveat for the probability conjecture, but it is not circular: the presentation theorem is not assumed in the moment theorem, and the random-group measure would still have the stated moments even if (a) failed for some K. Citations to [L WZB24], [Liu22], and [Liu25] are prior mathematical results with stated assumptions; although several are by the first author, the current paper proves its own imaginary-case extensions (Theorem 1.1, Theorem 1.2) rather than importing the target conclusion as an input. No specific equation or construction reduces to its own output, so no circularity is established.

Assumptions & free parameters 0 free parameters · 7 assumptions · 0 invented entities

The theory has no fitted constants. The moment value 1/[H^{Γ∞}:H^Γ] is derived from a point count, not tuned; the random group uses Haar measures without adjustable parameters. The random group X_{Γ,Γ∞} is a fully specified mathematical construction, not a new physical entity; no particles, forces, or unexplained degrees of freedom are introduced.

assumptions (7)
  • ad hoc to paper G^#_∅(K)_C is finitely admissibly generated (Theorem 1.2 condition (a))
    Assumed to construct the presentation; §5.2 states that finite generation of G^∞_∅(K) is an open question in general.
  • domain assumption Condition (b): primes ℓ dividing orders in C satisfy ℓ ∤ |Γ|, ℓ ≠ char(K), and µ_ℓ ⊄ K
    Restricts the scope of Theorem 1.2 and the probability conjecture; the authors conjecture this holds for 100% of K when µ_Q is trivial (Remark 7.2(2)).
  • domain assumption q ≡ 1 mod |Γ∞| for nonempty E_{Γ,Γ∞}(q^n, F_q(t))
    Needed for the existence of (Γ,Γ∞)-extensions of F_q(t) by local class field theory (§1.1).
  • standard math Homological stability of Hurwitz spaces (Landesman-Levy [LL25])
    Used in §4.2 to bound error terms in the trace formula and prove the fixed-q moment statement (1.2); cited as an external preprint.
  • standard math Behrend's trace formula for algebraic stacks
    The main counting tool for F_q-points of Hurwitz stacks (§3).
  • standard math Sawin's moment problem theorem [Saw20]
    Used in Theorem 6.3(4) to guarantee uniqueness of the probability measure from its moments.
  • standard math Schur-Zassenhaus theorem
    Used throughout to define the Γ-action on G^#_∅(K) and to count splittings (§1.1, Lemma 4.1).

how reviews work

0 comments
Cite this review

Pith. "Pith review of The imaginary case of the nonabelian Cohen--Lenstra heuristics." pith.science (2026). https://pith.science/paper/PRJPILZ3

@misc{pith2026250721558,
  author       = {Pith},
  title        = {Pith review of: The imaginary case of the nonabelian Cohen--Lenstra heuristics},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PRJPILZ3}},
  note         = {Machine review of arXiv:2507.21558}
}
abstract

For a finite group $\Gamma$, we study the distribution of the Galois group $G_{\emptyset}^{\#}(K)$ of the maximal unramified extension of $K$ that is split completely at $\infty$ and has degree prime to $|\Gamma|$ and $\textit{Char}(K)$, as $K$ varies over imaginary $\Gamma$-extensions of $\mathbb{Q}$ or $\mathbb{F}_q(t)$. In the function field case, we compute the moments of the distribution of $G_{\emptyset}^{\#}(K)$ by counting points on Hurwitz stacks. In order to understand the probability of the distribution, we prove that $G_{\emptyset}^{\#}(K)$ admits presentations of a specific form, then use this presentation to build random groups to simulate the behavior of $G_{\emptyset}^{\#}(K)$, and make the conjecture to predict the distribution using the probability measures of these random groups. Our results provide the imaginary analog of the work of Wood, Zureick-Brown, and the first author on the nonabelian Cohen--Lenstra heuristics.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

5 extracted references · 4 canonical work pages

  1. [1]

    Generalizing the GAGA Principle

    [BBCL22] Renee Bell, Jeremy Booher, William Y. Chen, and Yuan Liu, Tamely ramified covers of the projective line with alternating and symmetric monodromy , Algebra Number Theory 16 (2022), no. 2, 393–446. MR4412578 [BBH17] Nigel Boston, Michael R. Bush, and Farshid Hajir, Heuristics for p-class towers of imaginary quadratic fields, Math. Ann. 368 (2017), ...

    work page Pith review arXiv 2022

  2. [13]

    MR4240808 [WW21] Weitong Wang and Melanie Matchett Wood, Moments and interpretations of the Cohen-Lenstra-Martinet heuristics, Comment. Math. Helv. 96 (2021), no. 2, 339–387. MR4277275 Department of Mathematics, University of Illinois at Urbana-Champaign, 1409 W Green St, Ur- bana, IL 48109 USA Email address : yyyliu@illinois.edu Department of Mathematics...

    work page 2021

  3. [2008]

    Algebraic Geom

    MR2392026 [Ols15] Martin Olsson, Fujiwara’s theorem for equivariant correspondences, J. Algebraic Geom. 24 (2015), no. 3, 401–497. MR3344762 30 [Ols16] , Algebraic spaces and stacks , American Mathematical Society Colloquium Publications, vol. 62, American Mathematical Society, Providence, RI,

    work page 2015

  4. [2016]

    MR3495343 [R W06] Matthieu Romagny and Stefan Wewers, Hurwitz spaces, Groupes de Galois arithm´ etiques et diff´ erentiels, 2006, pp. 313–341. MR2316356 [Saw20] Will Sawin, Identifying measures on non-abelian groups and modules by their moments via reduction to a local problem (2020). preprint, arXiv:2006.04934v3. [Smi22] Alexander Smith, The distribution...

    work page Pith review arXiv 2020

  5. [2024]

    preprint, arXiv:2210.06279v2

    [SW22] Will Sawin and Melanie Matchett Wood, The moment problem for random objects in a category (2022). preprint, arXiv:2210.06279v2. [SW23] , Conjectures for distributions of class groups of extensions of number fields containing roots of unity (2023). preprint, arXiv:2301.00791v2. [Woo21] Melanie Matchett Wood, An algebraic lifting invariant of Ellenbe...

    arXiv 2022

Pith tools

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