Pith.

Number Theory · arXiv:2610.15257 v1 · submitted 11 Oct 2026 · 14 references

Post-critically finite polynomials over cyclotomic closures of number fields

Patrick Ingram, Jorge Mello

The paper says

“We prove that, for any number field K and and d at least 2, there are at most finitely many monic, centred, post-critically finite polynomials of degree d with coefficients in the extension of K by all roots of unity.”

from the abstract

Pith says

Machine review · deepseek-v4-flash · 14 Oct 2026 · what was read was not recorded · not checked by a person

It can be wrong. Challenge this review or reply to it.

Holds up, with conditions. The finiteness theorem is new and the main argument mostly checks out, but Lemma 6's super-attracting parameter is not justified as written, and that repair is needed before the contradiction in Lemma 9 can stand. This line is wrong

This review says the paper's mathematics is wrong. Pith did not check that against the paper's source.

1 major objection · 5 minor

How often objections from reviews like this one are right: not yet measured

Are you an author? What you can do.

Not checked: figures, code, what the 14 references say.

The paper says, at §5, Lemma 6 (and its use in Lemma 7 and Lemma 9)

The review did not quote this passage.

Pith says: Objection 1 of 1 · major

The super-attracting parameter is the hinge of the proof. Lemma 7's bound ord_x(W_{q(ℓ+1),qℓ}) ≥ ℓ+1 and the final contradiction in Lemma 9 (m'−1 ≥ ord_x(L) ≥ ord_y(W) ≥ m'+1) both require a parameter y at which a critical point c of infinite generic orbit becomes periodic modulo a point.

more of objection 1

The proof of Lemma 6, however, does not currently establish this from the cited results. In the r=1 case it infers from non-isotriviality and non-persistent preperiodicity that the pair (f,c) is 'not passive', citing [4, Theorem 2.5], and then applies [4, Proposition 2.4] to obtain a parameter at which c is periodic. The paper neither states the definition of passivity used in [4] nor verifies the exact hypotheses of these results. Under the standard local activity/passivity dichotomy the inference is not automatic: in z^2+c restricted to the main cardioid, the critical point 0 is passive, the family is non-isotrivial, and 0 is not persistently preperiodic. If the result in [4] that the authors need carries an algebraic or Zariski-open hypothesis that rules out such examples, that hypothesis must be stated and checked for the affine curve B and for the higher-dimensional reduction via the β-tuple pullback and the cover X → G_m^r. In addition, the text must address whether [4, Proposition 2.4] gives a periodic critical point or only a preperiodic one; Lemma 7's proof uses \(\bar g'(\bar c)=0\), which holds only when the critical point itself is periodic. If only preperiodic parameters are available, the high-order vanishing argument has no anchor. The phrase 'in the interest of avoiding some technicalities' in the proof of Lemma 6 marks exactly the place where this verification is omitted.

This review says the paper's mathematics is wrong. Pith did not check that against the paper's source.

5 minor objections

The paper says, at Abstract and Introduction

Pith says: minor objection 1 of 5

There is a duplicated word in the abstract: 'for any number field K and and d ≥ 2' should read 'and d ≥ 2'.

The paper says, at §5, Lemma 6

Pith says: minor objection 2 of 5

The sentence 'in the interest of avoiding some technicalities' should be replaced by a precise verification that the curve B and the β-pullback family satisfy the hypotheses of [4]; as written, the reduction from G_m^r to the one-dimensional case is a claim rather than a proof.

show 3 more

The paper says, at §6, Lemma 9

Pith says: minor objection 3 of 5

The symbol y is used both for the super-attracting parameter and for the coordinates of A^{m'} in equation (10); this makes the paragraph harder to follow and should be changed, for example by calling the parameter x and the coordinates y_i.

The paper says, at §5, Lemma 6

Pith says: minor objection 4 of 5

The assertion that, for monic centred polynomials, being isotrivial is the same as having constant coefficients is used without proof; it follows from the uniqueness of the monic centred normalization up to z ↦ ζz with ζ^{d−1}=1, and a one-sentence justification would be helpful.

The paper says, at Various

Pith says: minor objection 5 of 5

There are minor typographical errors: 'for all by finitely many v' should be 'for all but finitely many v' in §3; 'Then exists a linear form' should be 'Then there exists a linear form' in Lemma 4; and reference [13] should list 'A. Ostafe and M. Young' rather than 'A. Ostafe, Alina and M. Young'.

What would settle it

Pull the exact statements of [4, Theorem 2.5 and Proposition 2.4] and the definition of passivity used there. Verify: (i) Theorem 2.5 really has the contrapositive used here, with no hidden hypothesis such as the marked critical point belonging to the Julia set or not escaping; and (ii) Proposition 2.4 concludes periodic, not merely preperiodic, from that hypothesis. Then test the inference on f_c(z)=z^2+c over the main cardioid, where the usual activity dichotomy makes 0 passive while the family is non-isotrivial and 0 is not persistently preperiodic. If [4] uses a nonstandard notion that excludes this example, Lemma 6 should state that notion and verify it for the constructed finite cover B; if not, Lemma 6 is unsupported and Theorem 1 needs another argument.

from the skeptic's reading

How the review went

4 readings by deepseek-v4-flash · 14 Oct 2026

  1. Reader holds up, with conditions confidence moderate
  2. Skeptic partly agrees 11 min 9 s · 14,292 tokens read, 57,587 written
  3. Editor wrote the verdict above 2 min 3 s · 15,261 tokens read, 17,145 written
  4. Referee major revision: 1 major objection, 5 minor 8 min 57 s · 14,612 tokens read, 43,562 written

Not run on this paper: the simulated authors' defence.

Pith did not check this review's objections against the paper.

The review in full

The review in brief, machine-written

Read this one carefully before endorsing it: the theorem is new and the core argument is mostly sound, but Lemma 6, which supplies the super-attracting parameter, is not proved as written. The stress-test note is right about that step. My guess is the gap is fixable and the theorem is true, but a referee should not wave the lemma through.

What is actually new: Theorem 1. Neither Ingram's height bound (finite extensions only) nor the unicritical totally real case gives it, and the structure — Dvornicich–Zannier linear forms, Laurent's theorem, a multivariable Hajos lemma, and a dimension descent on exceptional families — is a real extension of the toolbox. The corollary on non-isotrivial families over B(K^cyc) is natural and correctly scoped, and they note the two-critical-orbit case already works over C.

What I checked and found sound: Lemma 5's Hajos proof, both the one-variable induction and the reduction to it by a generic s-tuple; Lemma 7's resultant identity (8) and the induction giving ord_y(Wq(ell+1),qell) >= ell+1 (the step where fq has f' as a factor is correct); Proposition 3's adelic bounds and escape argument; and the final contradiction in Lemma 9, where the pulled-back inequality plus Hajos' m'-1 against Lemma 7's m'+1 does close. The proof of Theorem 1 from Lemma 9 is clean, including Lemma 8's trimming of zero coordinates.

The soft spot is Lemma 6, and it is load-bearing. The implication used — non-isotrivial and not persistently preperiodic implies the marked critical point is not passive — is not true as a local statement: fc(z)=z2+c over the main cardioid is non-isotrivial, 0 is not persistently preperiodic, and 0 is passive with the orbit converging to an attracting fixed point. What might save the conclusion is global structure the paper does not use: B is affine and finite over Gm, so coefficient functions are nonconstant on a punctured curve and the family must be active somewhere. The authors' own "in the interest of avoiding some technicalities" marks the exact spot. Since Lemma 7's high-order vanishing, and therefore the contradiction in Lemma 9, is anchored at this parameter, the proof as written does not carry the load. This is the difference between accept and conditional.

Minor: the Lean formalization is relative and carries no commit hash, so it cannot be verified from the paper; the authors say so in advance, so it is a caveat, not a defect. Non-effectiveness is acknowledged.

Who this is for: arithmetic dynamicists and people working on unlikely intersections in dynamics; it finishes the line from Ingram's IMRN paper and pairs naturally with Zhang's abelian-extension result. Recommendation: send to a serious referee, and require a rewritten Lemma 6 with the exact Dujardin–Favre and DeMarco statements and a verified check of their hypotheses after the beta-tuple pullback. If that comes back, the rest of the paper holds.

The strongest objection

The proof's pivot is Lemma 6, which is the only source of the super-attracting parameter used in Lemma 7 to get ord_y(W_{q(m'+1), qm'}) >= m'+1, and hence the final contradiction with Hajos' bound ord_x(L) <= m'-1 in Lemma 9 (sentences b123 and b180). In the r=1 case, Lemma 6 relies entirely on [4, Theorem 2.5 and Proposition 2.4] through the inference: non-isotrivial and not persistently preperiodic implies not passive, and not passive yields a parameter at which the marked critical point is periodic. This inference is not automatic under the standard activity/passivity dichotomy for marked critical points. For example, in the family f_c(z) = z^2+c restricted to the main cardioid, the critical point 0 has a bounded orbit converging to an attracting fixed point, so it is passive in the usual DF/DeMarco sense; the family is non-isotrivial and 0 is not persistently preperiodic. The paper neither states the definition of passivity used in [4] nor checks hypotheses that might exclude this situation, such as requiring the marked critical point to lie in the Julia set or to be non-escaping. The higher-dimensional reduction via the beta-tuple pullback, the cover X -> G_m^r, and the curve B is plausible but inherits the same gap. If Lemma 6 fails, Lemma 7 has no super-attracting parameter at which to anchor the high-order vanishing, and the contradiction in Lemma 9 collapses.

I read the main computation: Proposition 3's height bounds, Lemma 4's Dvornicich-Zannier representation, Lemma 5's Hajos bound, and Lemma 7's resultant identity plus induction all check out internally. The reader's weakest-assumption analysis correctly identifies Lemma 6 as the structurally separate premise on which the contradiction rests. I partially agree: rather than doubting only the non-isotriviality of the beta-tuple pullback, I think the sharper problem is the paper's inference from non-isotriviality and non-persistent preperiodicity to non-passivity. Under the standard activity/passivity dichotomy this inference is false, since z^2+c over the main cardioid has a passive critical point, a non-isotrivial family, and no persistently preperiodic critical point. That means the appeal to [4] must be checked against the precise definition and hypotheses in that paper. If the cited theorem uses a different definition of passivity, Lemma 6 may still be true, but the current text does not supply the missing verification. The relative Lean formalization is also not independently checkable without a commit hash, but that is secondary; the mathematical gap is Lemma 6. I do not see a red flag in the internal algebra or in the induction over torus dimension, and the self-cited input [8] is a genuine published theorem used within its scope. Since the reader already rendered a CONDITIONAL verdict, my more specific concern sharpens the condition but does not by itself establish that Theorem 1 is false; the verdict should remain CONDITIONAL as given.

The referee's report

If the proof is correct, the result is a substantial arithmetic-dynamical analogue of Ribet's finiteness theorem for torsion points over cyclotomic closures: it gives finiteness of PCF conjugacy classes of fixed degree over K^cyc and, as stated in the introduction, finiteness of PCF specializations of non-isotrivial affine families at points of B(K^cyc). The paper is also commendably explicit about its hypotheses and limitations: it states that the proof is ineffective, it acknowledges the difficulties in the naive proof sketch, and it reports a relative Lean formalization whose axioms are the cited results. The internal algebra after Lemma 6 — the resultant identity (8), the Hajós induction, the truncation argument in Lemma 9, and the order-of-vanishing contradiction — is coherent and largely machine-checked. The decisive unresolved point is Lemma 6, the existence of a super-attracting parameter; because Lemmas 7 and 9 both lean on it, that lemma is the one load-bearing point that needs to be made fully rigorous.

Pith's reading of the paper (a machine's summary, not the authors')

This paper proves that, for any number field $K$ and any degree $d \geq 2$, only finitely many monic centred post-critically finite (PCF) polynomials of degree $d$ have all coefficients in $K^{\mathrm{cyc}}$, the cyclotomic closure of $K$ obtained by adjoining every root of unity. The result is the arithmetic-dynamics analogue of the finiteness of torsion points of abelian varieties over cyclotomic closures: PCF maps play the role of torsion points, and $K^{\mathrm{cyc}}$ is the same infinite extension. The proof is by contradiction and gives no effective bounds: infinitely many such polynomials would force the existence of a parametrised family over a torus with infinitely many torsion points specializing to PCF maps, and the authors show such families cannot exist by pitting a combinatorial bound on the order of vanishing of a resultant against a bound coming from the dynamics of a super-attracting parameter. A corollary states that any non-isotrivial affine family of monic polynomials has only finitely many PCF specializations at points of the cyclotomic closure of the base field.

What carries the argument. The argument runs on a conflict between two bounds on the order of vanishing of a single resultant $W_{\ell,k}(u) = \prod (f^\ell(c) - f^k(c))$, the product taken over critical points with infinite forward orbit, viewed as a Laurent polynomial in the parameter $u$ on a torus $G_m^r$. Hajós' Lemma says a nonzero Laurent polynomial expressible with at most $T$ monomials vanishes to order at most $T-1$ at any point. On the other side, a super-attracting parameter — a torsion point of the parameter torus where a critical point with infinite forward orbit becomes periodic — forces $W_{q(k+1),qk}$ to vanish to order at least $k+1$ (Lemma 7), and Lemma 6, imported from [4], guarantees such a parameter for every non-isotrivial family. The contradiction is built on an exceptional family: infinite PCF data over $K^{\mathrm{cyc}}$ is compressed — via the linear-form representation of [5] and the theorem on closures of infinite torsion sets in tori [10] — onto a torsion coset where $W$ is a lacunary Laurent polynomial with a bounded number of terms; the two bounds then clash, since $m'-1 \geq \mathrm{ord}_x(L) \geq m'+1$.

Core claim. The paper's central claim is Theorem 1: fix a number field $K$ and $d \geq 2$; there are only finitely many monic centred post-critically finite (PCF) polynomials of degree $d$ with all coefficients in $K^{\mathrm{cyc}}$, the extension of $K$ by all roots of unity. Equivalently — since two monic centred polynomials are conjugate only by $z \mapsto \zeta z$ with $\zeta^{d-1} = 1$ — only finitely many conjugacy classes of monic PCF polynomials of degree $d$ are defined over $K^{\mathrm{cyc}}$. A corollary the authors draw: if $f(z) \in K[B,z]$ is a non-isotrivial family of monic polynomials over an affine curve $B$, then only finitely many specializations $f_t$ at $t \in B(K^{\mathrm{cyc}})$ are PCF; this is contrasted with one-parameter families such as $z^d + c$, which have infinitely many PCF specializations over $\mathbb{Q}$. The proof of the theorem establishes the stronger intermediate statement that no exceptional family exists: no non-isotrivial family of monic centred polynomials over a positive-dimensional torus admits infinitely many torsion parameters at which the specializations are distinct PCF maps.

The premise it rests on. The argument's final contradiction leans entirely on Lemma 6, which asserts that every non-isotrivial monic centred family over a torus $G_m^r$ has a super-attracting parameter. That lemma is imported from the one-parameter results of [4] by a pullback the authors describe as done in the interest of avoiding some technicalities; if the curve reduction fails, or the cited hypotheses are not met in the stated generality, the final clash of orders collapses.

If the paper is right (Pith's inferences, not the authors')

Finiteness over the cyclotomic closure: for fixed $K$ and $d \geq 2$, only finitely many monic centred PCF polynomials, and hence only finitely many conjugacy classes of such maps, have coefficients in $K^{\mathrm{cyc}}$; this is the paper's Theorem 1.

Finiteness of PCF specializations in families: over any affine curve $B$ with $f \in K[B,z]$ non-isotrivial, only finitely many $t \in B(K^{\mathrm{cyc}})$ give PCF $f_t$.

The limitation to higher degree is real: families with a single symmetry class of infinite critical orbit, like $z^d + c$, still have infinitely many PCF specializations over $\mathbb{Q}$, so the theorem does not hold for arbitrary one-parameter families.

No effective control is obtained: the proof is non-constructive and yields no bound on the finite number of PCF polynomials for given $K$ and $d$.

The template of the proof — adelic bounds, linear forms in roots of unity, the torsion-coset theorem, and a clash between a Hajós-type combinatorial bound and a super-attracting dynamical bound — looks portable to the abelian closure $K^{\mathrm{ab}}$, the direction the authors say they explored before encountering independent work [14] on a Bogomolov property over abelian extensions; if the same

Because the proof is non-effective, the natural next step is to seek any quantitative control — even a uniform bound on the number of PCF maps in terms of $d$ and $[K:\mathbb{Q}]$ — or to pinpoint where effectivity is genuinely lost in the chain.

The authors' Lean formalization is explicitly relative, proving Theorem 1 conditional on the cited results; independent audit of the super-attracting-parameter lemma (Lemma 6) is therefore the highest-leverage check a reader can perform to gain full confidence in the theorem.

Check this review

Pith stored no record of this review's passes and took no fingerprint of it. Reviews made by code from 13 Nov 2026 on carry both.

How often reviews like this one are right

Reviews like this one: written by deepseek-v4-flash, where what was read was not recorded.

How often objections from reviews like this one are right: not yet measured

Every value above, in the review's public record (JSON)

Reviews by people

This review was written by a machine. No person has reviewed this paper on Pith yet. A scientist listed on Pith can review it for a fee, and their review is published here whatever it concludes.

Request a human review

How Pith checks papers · The full audit of this paper

The paper

Abstract

We prove that, for any number field K and and d at least 2, there are at most finitely many monic, centred, post-critically finite polynomials of degree d with coefficients in the extension of K by all roots of unity.

The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. No sentence in this record was written by a model.

abstract compiled · 2026-10-14

The authors' main claim, as the classifier picked it out 1.00

“We prove that, for any number field K and and d at least 2, there are at most finitely many monic, centred, post-critically finite polynomials of degree d with coefficients in the extension of K by all roots of unity.”

How the main claim is stated: asserted 1.00

What the abstract itself states, as probabilities: claims first 0.03 · states limitation 0.24 · human participants 0.01 · controlled comparison 0.01 · randomized assignment 0.01 · states quantitative result 0.42 · compares with prior results 0.04 · mentions code or data availability 0.01

The abstract, sentence by sentence

This table labels the abstract only (arXiv listing abstract v1). The number is the probability the classifier gave the label.

SentenceRoleThe authors' own workStates the main conclusionStrength
We prove that, for any number field K and and d at least 2, there are at most finitely many monic, centred, post-critically finite polynomials of degree d with coefficients in the extension of K by all roots of unity. result 0.94 0.94 0.94 asserted 1.00

What kind of paper the classifier reports this is: theoretical result 1.00 · evidence from proof or derivation 1.00

What was examined, and what was not

Sources read

  • arXiv listing abstract v1

Not examined

  • full text
  • figures and tables
  • references and citing sentences
  • relations to other papers

Levels run

  • P abstract

Not run

  • S sections
  • T body sentences
  • C citations
  • R cross-paper relations

How the labels were made

Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-10-14: 1 request, 1,683 input tokens.

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. This record as JSON

References (14)
  1. [4] R. Dujardin and C. Favre,Distribution of rational maps with a preperiodic critical point, Amer. J. Math.130(2008), no. 4, 979–1032 on Pith
  2. [1] B. Branner and J. H. Hubbard,The iteration of cubic polynomials. Part I: The global topology of parameter space, Acta Math.160(1988), 143–206 on Pith
  3. [2] E. Chen,Avoiding algebraic integers of bounded house in orbits of rational functions over cyclotomic closures, Proc. Am. Math. Soc.146(10) (2018) 4189–4198 on Pith
  4. [3] L. DeMarco,Bifurcations, intersections, and heights, Algebra Number Theory10(2016), no. 5, 1031–1056 on Pith
  5. [5] R. Dvornicich and U. Zannier,Cyclotomic diophantine problems (Hilbert irreducibility and invariant sets for polynomial maps), Duke Math. J.139(2007), no. 3, 527–554 on Pith
  6. [6] C. Favre and T. Gauthier, The arithmetic of polynomial dynamical pairs,Ann. of Math. Stud.214Princeton University Press, Princeton, NJ, 2022, xvii+232 pp on Pith
  7. [7] A. Ferraguti, A. Ostafe, and U. Zannier,Cyclotomic and abelian points in backward orbits of rational functions, Adv. Math.438(2024) Paper No. 109463, 27 pp on Pith
  8. [8] P. Ingram,A finiteness result for post-critically finite polynomials, Int. Math. Res. Not. (2012), no. 3, 524–543 on Pith
  9. [9] N. M. Katz and S. Lang,Finiteness theorems in geometric classfield theory, With an ap- pendix by Kenneth A. Ribet, Enseign. Math. (2)27(1981), no. 3-4, 285–319 (1982) on Pith
  10. [10] M. Laurent, ´Equations diophantiennes exponentielles, Invent. Math.78(1984), no. 2, 299– 327 on Pith
  11. [11] J. H. Loxton,On the maximum modulus of cyclotomic integers, Acta Arith.22(1972), 69–85 on Pith
  12. [12] C. Noytaptim and C. Petsche,Totally real algebraic integers in short intervals, Jacobi poly- nomials, and unicritical families in arithmetic dynamics, Bull. Lond. Math. Soc.56(2024), no. 1, 364–379 on Pith
  13. [13] A. Ostafe, Alina and M. Young,On algebraic integers of bounded house and preperiodicity in polynomial semigroup dynamics, Trans. Amer. Math. Soc.373(2020), no. 3, 2191–2206 on Pith
  14. [14] G.-R. Zhang,A Bogomolov property for moduli spaces of polynomials over abelian exten- sions,arXiv:2610.06382 on PithPith's reviewarXiv
Cited by 1 Pith paper

Reviewed papers on Pith that cite this one. The full citation record

  1. Postcritically Finite Unicritical Polynomials over CM Fields math.NT · Nov 2026

Challenges and discussion

Disagree with the review, or know something it missed? Leave a comment, or challenge this review formally: a challenge becomes a permanent entry in this paper's record, and its decision is published here.

No comments yet.

Sign in to comment