{"id":"fcae4fad-386a-4573-811f-3c41fdc6d2fb","arxiv_id":"2610.06382","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"For every number field K and degree d ≥ 2, the K^ab-rational postcritically finite conjugacy classes of degree-d polynomials form a finite set.","lead":"The paper proves that for every number field K and every degree d ≥ 2, only finitely many K^ab-rational points of the degree-d polynomial moduli space can be postcritically finite: small critical height forces zero critical height. It is the first Bogomolov-type finiteness result for polynomial moduli over maximal abelian extensions, beyond the reach of classical Northcott arguments.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek/deepseek-v4-flash","headline":"Proposition 3.2's forcing of the special fiber into Gamma_e rests on unproved flatness and positivity of the model weights; if one atom of the Chambert-Loir measure lies off Gamma_e, both contradiction cases of Theorem 1.1 collapse.","rationale":"The reader's conditional verdict identifies Proposition 3.2 as the central fragile hinge, and I agree. The theorem's two final contradictions are both built on Proposition 3.2: the wild case needs the output correspondence to be bi-finite, and the tame case needs the numerical pullback relation π_2^*A_Q≡pπ_1^*A_Q. The forcing step from the limiting measure to the special fiber Γ_e depends on every atom of the Chambert-Loir measure lying in φ^{-1}(T_e(δ)), which in turn depends on the positivity of a_k(H^m·Y_k) and on the flatness/finiteness of the model W. These conditions are asserted but not proved in detail in the text; they are plausible and probably correct, but they are exactly the kind of routine verification that can hide a degenerate-special-fiber failure. I found no explicit counterexample or internal contradiction. The paper has substantial independent support: the main auxiliary constructions are proved in the text, the height formula is derived from explicit potentials, and the local inertia estimate is elementary and correctly stated. For these reasons my read does not change the reader's conditional verdict; a referee-level audit of Proposition 3.2's model-measure step is still warranted before acceptance.","tokens_in":33107,"tokens_out":56999,"duration_ms":745117,"concrete_test":"Independently re-derive the step (3.4)⇒(3.5) for the model W constructed in Proposition 3.2, without invoking Yuan's theorem as an unexplained black box: write out the model-measure atoms ξ_k and the weights a_k(H^m·Y_k), and verify that every component Y_k of W_κ satisfies Φκ(Y_k)⊆Γ_{e,κ} and a_k(H^m·Y_k)>0. Concretely, take X=P^1_{Z_p}, A=O(1), and choose a correspondence C whose closure over Z_p has a non-reduced or multi-component special fiber; compute the normalization W, the intersection numbers (H^m·Y_k), and the reductions of the divisorial points. If any special-fiber component is not contained in Γ_e, or if any weight vanishes, Proposition 3.2 fails and Theorem 1.1's wild and tame contradiction cases both lose their forcing mechanism.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The single most load-bearing point is the forcing step of Proposition 3.2, specifically the chain (3.3)-(3.5). For an irreducible component C of Z(B) dominating both factors, the proof selects a generic small-height sequence in ν^{-1}(B∩C) and uses Yuan's equidistribution to identify the limit measure on W^{an}_{Lv} with the Chambert-Loir model measure. The inclusion Supp(µ_v)⊆(φ^an)^{-1}(T_e(δ)) then forces φ^an(ξ_k)∈T_e(δ) for every atom ξ_k, and hence forces each component Y_k of the special fiber into Γ_e. The step is valid only if every atom has positive mass: the model measure formula weights atoms by a_k(H^m·Y_k), and the proof asserts positivity because H is relatively ample on a normal projective flat R-model W. This hides two conditions: (i) the normalized closure W of C in X_R×X_R is flat over R; (ii) Φ=(Π1,Π2) is finite, so H=Φ^*(A_R⊠A_R) restricts with positive degree to every m-dimensional component of W_κ. If W has an embedded or vertical component, or if some intersection number vanishes, the corresponding atom is unconstrained and (3.5) fails. The remainder of Theorem 1.1 depends on (3.5): the wild case only needs the resulting bi-finite correspondence, and the tame case needs the numerical relation π_2^*A_Q≡pπ_1^*A_Q that Lemma 3.3 contradicts. The proof also uses Galois invariance of B to keep every point of the approximating orbits inside T_e(δ); after base change to L' this must be checked for the chosen lifts z_j. These are audit points rather than observed contradictions, but they are load-bearing.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek/deepseek-v4-flash","summary":"The paper proves Theorem 1.1: for every integer d ≥ 2 and every number field K there exists ε_{K,d} > 0 such that the set of K^ab-rational points of MPoly^d with critical height below ε_{K,d} coincides with the finite zero-height locus, i.e., the set of postcritically finite classes; the critical height satisfies the strong Bogomolov property over abelian extensions. The proof combines: a universal critical-divisor family (X_d, F_d, p_d, s_d) with α PCF iff s_d(α) is F_d-preperiodic (Prop. 2.1); a nef adelic line bundle A with the exact orbit-height formula h_A(F_d^n s_d(α)) = k_d d^n h_crit^max(α), k_d = (d−1)lcm(1,…,d) (Thm. 2.2); the relative-degree formula deg G = d^{sl} (Prop. 2.3); the Noetherian orbit-closure dichotomy with Thurston rigidity (Lemma 4.1, Prop. 4.2); the intertwined relation ≈ with bounded classes (Prop. 3.1, [FG22]); a small-point correspondence inside diagonal/Frobenius tubes (Prop. 3.2); and p-adic inertia estimates (Lemma 3.4). In the wild case, order-p inertia elements yield p pairwise distinct intertwined conjugate points, contradicting p > M_d; in the tame case, the Frobenius tube yields π_2^*A_Q ≡ pπ_1^*A_Q while Lemma 4.3 gives π_1^*U = π_2^*U, contradicting Lemma 3.3. The main auxiliary results are proved in the text, and the contradiction machinery does not re-use the theorem's conclusion.","tokens_in":33380,"tokens_out":40952,"duration_ms":585070,"significance":"If correct, Theorem 1.1 is a substantial result: it gives the strong Bogomolov property for the critical height on polynomial moduli over the maximal abelian extension, going decisively beyond Northcott-based arguments, which only bound the residue degree, and it yields finiteness of PCF classes over abelian extensions. The paper's architecture is coherent and contains genuine new machinery: the universal critical-divisor family, the explicit exact orbit-height formula with parameter-free constants, the relative-degree computation, the honest local inertia estimates (Lemma 3.4), and the intertwing/small-point correspondence setup. The auxiliary propositions (2.3, 4.2, Lemma 3.3, Lemma 4.3 conditional on (1)) are largely self-contained, which is a strength. The central weakness is that the load-bearing forcing step of Proposition 3.2, which underpins both contradiction cases, appears circular as written, and the wild case's Galois-invariance claim is not justified. These are repairable in principle, so I do not judge the claim false, but the gaps are genuine and central.","major_comments":[{"comment":"The forcing step rests on the assertion that every weight a_k(H^m·Y_k) in the Chambert–Loir model measure is positive 'because H is relatively ample'. Flatness of the model is not the issue: W, being the normalization of the scheme-theoretic closure of an irreducible generic fiber over a DVR, is torsion-free and hence flat over R, and normalization kills embedded components. The real problem is the relative ampleness of H = Π_1^*A_R ⊗ Π_2^*A_R. On the special fiber, for a curve C ⊆ W_κ one has H·C = deg(Π_1^*A_κ|_C) + deg(Π_2^*A_κ|_C), which is positive for every C if and only if no curve is contracted by both Π_1 and Π_2, i.e., if and only if Φ_κ = (Π_1,Π_2)_κ is quasi-finite. That quasi-finiteness is precisely the finiteness conclusion the proposition derives from (3.5). Thus the positivity of all atoms used to force Φ_κ(W_0) ⊆ Γ_{e,κ} is equivalent to the conclusion being proved; the derivation of finiteness of π_i and of the numerical relation (3.2) is circular at this point. The same issue affects the generic fiber: H = π_1^*A ⊗ π_2^*A is only shown 'nef and ample' if (π_1,π_2) is finite, which is again the proposition's conclusion. Because (3.5) drives both contradiction cases of Theorem 1.1 (the wild case needs only the resulting bi-finite correspondence, the tame case needs the relation π_2^*A_Q ≡ pπ_1^*A_Q), this is a load-bearing gap. The repair should supply an independent argument for the required relative ampleness or bigness of H, possibly by following [JSX26, §6] more closely, or should otherwise constrain the components of W_κ with vanishing weight.","section":"§3.2, Proposition 3.2 (chain (3.3)–(3.5))","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The claim that 'Since Gal(L^ab/L) is abelian, BIn is Gal(L^ab/L)-invariant' is not justified as written. For γ ∈ Gal(L^ab/L), abelianness gives γ τ_{L(x)} γ^{-1} = τ_{L(x)}, so γ(x, τ_{L(x)}x) = (γx, τ_{L(x)}γx). This point lies in BIn = {(y, τ_{L(y)}y)} only if τ_{L(x)} agrees on L(γx) with the element τ_{L(γx)} chosen for the field L(γx). Since for each field E the element τ_E is fixed arbitrarily, nothing forces this compatibility across the Galois orbit of x; indeed one can choose the τ_E's to fail it. The Galois-invariance of B is a hypothesis of Proposition 3.2 (needed for the orbit measures µ_j and for the tube containment on all conjugates), so without a consistency argument the wild case does not currently fall under Proposition 3.2. The tame case is clean in this respect because a single σ is used for all x. A repair would consist of a systematic Galois-equivariant choice of the τ_E's, or of working with the Galois orbit of BIn after proving the tube inclusion for all conjugates.","section":"§4.3, Case 1 (wild case), definition of BIn","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The sentence 'it is clear that there is a generic sequence (z_j) ⊆ ν^{-1}(B∩C) in W/Q with h_H(z_j) < 1/j' is too quick: a rational point of C need not lift to a rational point of the normalization W, since ν is finite birational and can be non-trivial over singular points. A few sentences explaining that the lift can be chosen over Q̄, that the Galois orbits of the lifts remain in ν^{-1}(B∩C) (because B and T_e(δ) are Gal(Q̄/L)-stable and heights are Galois-invariant), and that small points of B are dense in C by (3.3), would make the equidistribution step airtight.","section":"§3.2, Proposition 3.2","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The application of [GN17, Proposition 5.3(i)] to conclude W = π_1(W) × π_2(W) when the two polynomials are not intertwined would benefit from an explicit statement and verification of that proposition's stated hypotheses (non-exceptional polynomial components, invariant subvariety inside the product affine space, and the dichotomy producing the product decomposition). This is an audit request, not a demand to reprove the cited result; a one-line check would remove an otherwise opaque step.","section":"§4.2, Lemma 4.3(1)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The symbol H is used for three different objects: the subgroup H = Gal(F_m/F_{m−1}) in the proof of Lemma 3.4(1), the hyperplane H∞ in the proof of Proposition 2.3, and the adelic line bundle H = π_1^*A ⊗ π_2^*A on W in Proposition 3.2. Please disambiguate at least two of these to reduce reader confusion.","section":"Notation","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The phrase 'red is the reduction for the proper model X ×_{Z_p} X' appears before the model W of Proposition 3.2 is introduced; aligning the notation with the later X_R and stating that T_e(δ) is defined on (X_{L_v} × X_{L_v})^{an} would improve clarity. This is purely presentational.","section":"§3.2, tubes","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"major_revision"},"author_rebuttal":null,"desk_editor":{"model":"deepseek/deepseek-v4-flash","letter":"The headline is simple: this is a real result, not an incremental one. Theorem 1.1 — strong Bogomolov for the critical height on MPoly^d over K^ab, giving finiteness of PCF classes over that infinite extension — is new. Ingram's height comparison only reaches bounded-degree extensions, and the cyclotomic case was handled separately. If it survives, it is a substantial advance in arithmetic dynamics.\n\nWhat the paper does well: nothing load-bearing is cited from the air. The universal critical-divisor family (Prop 2.1), the nef adelic bundle with the exact orbit-height formula (Thm 2.2), the relative-degree computation (Prop 2.3), and the intertwined-relation machinery are all built out in the text. The proof is honestly split into wild and tame cases, and the authors state their own limitation that h_max^crit is a sum of local maxima rather than the maximum of the global canonical heights, then bridge it with the bound (2.3). The orbit-height identity holds by construction of A, so there is no pretense that the height comparison is an independent prediction.\n\nThe soft spots, in proportion. The forcing step in Proposition 3.2 is the place I would send a referee to recompute, and it is the one genuinely delicate hinge. But part of the stress-test worry does not land: the model W is the normalization of a closure over a DVR, and integral schemes over DVRs are flat, so the flatness point resolves itself. The positivity of the weights a_k(H^m · Y_k) is correct for top-dimensional components, which is all the model measure needs; the sentence asserting it for every k is loose because it silently requires dim Y_k = m, but that looseness is benign. The real audit points are elsewhere: the hypothesis check for [GN17, Proposition 5.3(i)] inside Lemma 4.3(1), and two load-bearing citations that are unpublished preprints — [JSX26] and [Li24]. A referee should confirm the split-polynomial endomorphism result and the norm-metric machinery are used inside their stated hypotheses, not just that they appear in the bibliography.\n\nOn the AI disclosure: it is unusually honest, but the claim that all AI-assisted material was independently verified is not checkable. The practical consequence is that Theorem 2.2 and Proposition 3.2 should be treated as if written by a student with an unstated assistant — recompute, do not assume.\n\nThis paper deserves a serious referee. It is important enough and the architecture is sound enough that desk rejection would be wrong. I would want the referee to spend their time on §3.2 and Lemma 4.3, and to read the two unpublished preprints before signing off. For my part I would cite this if the main theorem holds up, and I would bring it to a reading group of arithmetic dynamicists.","headline":"Genuinely new Bogomolov-type theorem over abelian extensions, coherently built; the load-bearing model-measure forcing in Proposition 3.2 is audit-grade rather than broken, and the unpublished citations plus the AI disclosure deserve referee attention.","tokens_in":34086,"tokens_out":3478,"duration_ms":24538},"pith_extraction":{"msc":["37P30","37P45","11R20","14G40","37P05","37P15"],"pacs":[],"model":"deepseek/deepseek-v4-flash","headline":"For every number field $K$ and degree $d\\geq 2$, the paper proves that the critical height satisfies the strong Bogomolov property on the polynomial moduli space $\\mathrm{MPoly}^d$ over the maximal abelian extension $K^{\\mathrm{ab}}$: a…","keywords":["critical heights","Bogomolov property","postcritically finite polynomials","polynomial moduli","abelian extensions","adelic line bundles","equidistribution","intertwined polynomials"],"falsifier":"A decisive check: for $d = 2$, $\\mathrm{MPoly}^2 \\simeq \\mathbb{A}^1$ and $h_{\\mathrm{crit}}(a) = \\hat h_{f_a}(0)$ for $f_a(z) = z^2 + a$, so a direct test is to search over $K^{\\mathrm{ab}}$-points, for example through ramified $p$-power towers over fixed primes, for any pairwise-distinct sequence of non-PCF parameters with $\\hat h_{f_a}(0) \\to 0$, which would refute Theorem 1.1 outright; alternatively, computing the Chambert-Loir model measure on a model whose special fiber contains a component with $(H^m \\cdot Y_k) = 0$ would pinpoint whether the descent in Proposition 3.2 fails before the theorem itself does.","tokens_in":34244,"feed_emoji":"🎯","tokens_out":13888,"duration_ms":100915,"temperature":0.7,"pith_summary":"The paper proves that for every number field $K$ and every degree $d \\geq 2$ there is a positive constant $\\epsilon_{K,d}$ with the following effect: a $K^{\\mathrm{ab}}$-rational point of the moduli space of degree-$d$ polynomials has critical height either exactly zero or at least $\\epsilon_{K,d}$. Since the critical height vanishes precisely on the postcritically finite (PCF) classes, this says only finitely many $K^{\\mathrm{ab}}$-rational points are PCF. The result matters because $K^{\\mathrm{ab}}$ is an infinite extension, where Northcott's theorem does not apply: the known finiteness of bounded-degree PCF classes, via comparability of the critical height with an ample height, has no handle on points whose field of definition grows without bound. The paper delivers finiteness over $K^{\\mathrm{ab}}$ anyway, in the strong form of a uniform height gap, by a new application of the small-points correspondence method of Ji\\textendash{}Song\\textendash{}Xie to Galois-invariant sets, combined with a uniform bound on intertwined classes and separate wild and tame ramification estimates.","feed_headline":"Finitely many PCF classes over abelian extensions","feed_subtitle":"A uniform positive constant separates zero critical height from positive values on degree-d moduli over K^ab","key_machinery":"The argument rests on four interlocking pieces. First, a universal critical-divisor family: an affine variety $X_d$ with a finite endomorphism $F_d$, a morphism $p_d : X_d \\to \\mathrm{MPoly}^d$, and a section $s_d$ such that $\\alpha$ is PCF exactly when $s_d(\\alpha)$ is $F_d$-preperiodic. Second, a nef adelic line bundle $\\overline{A}$ on a normal projective compactification of $X_d$ with the exact orbit-height identity $h_{\\overline{A}}(F_d^{\\circ n}(s_d(\\alpha))) = k_d d^n h_{\\mathrm{crit}}^{\\max}(\\alpha)$, where $k_d = (d-1)\\operatorname{lcm}(1,\\dots,d)$; this converts the critical height into the height of explicit orbit points on a variety whose models have controlled good fibers. Third, the intertwined relation $\\approx$ on $\\mathrm{MPoly}^d$, a constructible equivalence over $\\mathbb{Q}$ whose classes are uniformly bounded by a constant $M_d$; this bound is the target the contradiction aims to break. Fourth, the small-points correspondence, Proposition 3.2: for a Galois-invariant set of height-$<\\epsilon$ points lying inside a fixed closed diagonal or Frobenius tube $T_e(\\delta)$, Yuan's equidistribution drives the Chambert-Loir model measure into the special fiber $\\Gamma_e$, forcing every component dominating both factors to be bi-finite with the two pulled-back line bundles proportional by $e$; this structural fact makes both ramification cases contradictory. The relative-degree formula $\\deg G = d^{sl}$ (with $s = \\dim Y - \\dim V$) and the local ramification estimates of Lemma 3.4 (an order-$p$ inertia subgroup with displacement bound $p^{-1/(p-1)}$ in the wild case, and ramification index dividing $p-1$ in the tame case) close the two cases.","core_discovery":"The central claim, Theorem 1.1, is that on each moduli space $\\mathrm{MPoly}^d$ the critical height $h_{\\mathrm{crit}}$ obeys the strong Bogomolov property over the maximal abelian extension $K^{\\mathrm{ab}}$ of any number field: there exists $\\epsilon_{K,d} > 0$ such that $\\{\\alpha \\in \\mathrm{MPoly}^d(K^{\\mathrm{ab}}) : h_{\\mathrm{crit}}(\\alpha) < \\epsilon_{K,d}\\} = \\{\\alpha \\in \\mathrm{MPoly}^d(K^{\\mathrm{ab}}) : h_{\\mathrm{crit}}(\\alpha) = 0\\}$, a finite set. In other words the PCF locus over $K^{\\mathrm{ab}}$ is finite for each degree, and the positive values of the critical height on $K^{\\mathrm{ab}}$-points are bounded away from zero by a constant depending only on $K$ and $d$. The proof runs by contradiction: a hypothetical sequence of distinct $K^{\\mathrm{ab}}$-points with heights tending to zero is packaged into a generic algebraic family carrying a finite orbit-endomorphism and a nef adelic line bundle whose orbit heights are explicit powers of $d$ times the critical height; the sequence is then split into a wildly ramified and a tamely ramified part over a Chebotarev prime, and each part forces an impossible configuration of intertwined parameters, in the wild case $p$ pairwise-distinct intertwined points contradicting the uniform bound $M_d$, and in the tame case a Frobenius-scaled line-bundle proportionality that cannot coexist with a nonconstant function shared by the two projections.","pith_inferences":["The bounded-class intertwined relation looks like the portable invariant: wherever a moduli problem carries (i) an exact orbit-height encoding of the dynamical finiteness predicate and (ii) a constructible equivalence on parameters with uniformly bounded finite classes stable under the relevant Galois twists, the same correspondence argument should give the strong Bogomolov property over $K^{\\math","The abelian hypothesis enters at two essential places — commutativity of $\\mathrm{Gal}(L^{\\mathrm{ab}}/L)$ (used to keep the diagonal-tube sets $B_{In}$ Galois-invariant) and local Kronecker\\textendash{}Weber (which supplies the order-$p$ inertia subgroup for wildly ramified abelian extensions); an analogous theorem over a non-abelian infinite extension would need a new mechanism for at least one ","The proof is by contradiction and the gap constant is not made effective; a natural testable extension is to seek explicit lower bounds, which for $d=2$ becomes a concrete computation since $\\mathrm{MPoly}^2 \\simeq \\mathbb{A}^1$ and $h_{\\mathrm{crit}}(a) = \\hat h_{f_a}(0)$ for $f_a(z) = z^2 + a$."],"forward_implications":["Theorem 1.2: for every number field $K$ and degree $d\\geq 2$, the set of PCF classes in $\\mathrm{MPoly}^d(K^{\\mathrm{ab}})$ is finite — the motivating finiteness, now obtained where Northcott's theorem cannot reach.","Theorem 1.3: the same finiteness holds over the maximal cyclotomic extension $K^c = K(\\mu_\\infty)$, with a second, independent route sketched in Section 1.4 via the cyclotomic integral-point rigidity theorem of Ji\\textendash{}Xie\\textendash{}Zhang.","Because $h_{\\mathrm{crit}}$ is comparable to an ample Weil height on $\\mathrm{MPoly}^d$, the theorem constrains all small-height $K^{\\mathrm{ab}}$-points, not only PCF ones: below $\\epsilon_{K,d}$ there is nothing but the vanishing locus.","The paper proposes the analogous statements for the moduli space $\\mathcal{M}_d$ of degree-$d$ rational maps away from the flexible Latt\\`es locus (Conjectures 1.5\\textendash{}1.7), noting that the flexible Latt\\`es locus must be excluded because it contains infinite PCF families when $d$ is a perfect square."],"supporting_citations":[{"why":"Supplies the small-points correspondence method (Theorem 6.6 and the proof of Lemma 6.14) that Proposition 3.2 adapts to Galois-invariant sets lying in closed tubes.","marker":"[JSX26]"},{"why":"Supplies the uniform bound $M_d$ on intertwined classes and the constructibility of the intertwined relation over $\\mathbb{Q}$; $p > M_d$ is the contradiction target of the wild case.","marker":"[FG22]"},{"why":"Proposition 5.3(i), the invariant-subvariety dichotomy for split polynomial maps, is the split step in Lemma 4.3(1) forcing the generic point of the correspondence onto the intertwined relation.","marker":"[GN17]"},{"why":"Yuan's equidistribution theorem, together with the model measure, forces the special fiber into the diagonal or Frobenius graph in the descent of Proposition 3.2.","marker":"[Yua08]"},{"why":"Chambert-Loir model measures provide the discrete measure at divisorial points, positivity of whose weights is required for the small-point descent.","marker":"[CL06]"},{"why":"Provides the adelic-line-bundle framework (nefness, convergence of metric sequences, heights of subvarieties) used to construct and descend the line bundle $\\overline{A}$.","marker":"[YZ26]"},{"why":"Gives comparability of the critical height with an ample moduli height and the Nullstellensatz estimates (Lemma 2.4) used in the metric construction; it also supplies the Northcott baseline the paper bypasses.","marker":"[Ing12]"},{"why":"Supplies the Chebotarev density theorem and Kronecker\\textendash{}Weber (global and local), giving the split prime $p$ and the structure of abelian extensions behind the ramification dichotomy.","marker":"[Neu99]"},{"why":"Provides the local-field ramification facts behind the inertia displacement bound $|\\tau(z)-z|_p \\leq p^{-1/(p-1)}$ in Lemma 3.4.","marker":"[Ser79]"},{"why":"Theorem 2.2 therein supplies the Thurston rigidity used to rule out the degenerate case where every point of $V$ is PCF, forcing the orbit-closure dichotomy of Lemma 4.1.","marker":"[BIJ+14]"}],"fun_headline_variants":["Strong Bogomolov gap for critical height over abelian extensions","PCF classes over K^ab are finite for every degree","Critical height has uniform positive gap on K^ab moduli","Finitely many postcritically finite points over abelian extensions"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The load-bearing step is the correspondence descent of Proposition 3.2: a Galois-invariant family of points whose heights go to zero, lying inside a fixed closed diagonal or Frobenius tube, must be carried by components finite over both factors with the two pulled-back line bundles proportional by $e$, and this descent through the special fiber — via Yuan's equidistribution landing in the Chambert-Loir model measure — requires the model's special fiber to be non-degenerate with positive intersection weights on every component.","fun_headline_variants_meta":{"raw":{"variants":["Strong Bogomolov gap for critical height over abelian extensions","PCF classes over K^ab are finite for every degree","Critical height has uniform positive gap on K^ab moduli","Finitely many postcritically finite points over abelian extensions"]},"model":"deepseek/deepseek-v4-flash","effort":"low","cost_usd":0.000159,"raw_usage":{"cost":0.000159208,"is_byok":false,"cost_details":{"upstream_inference_cost":0.000159208,"upstream_inference_prompt_cost":0.000123368,"upstream_inference_completions_cost":3.584e-05},"total_tokens":1214,"prompt_tokens":1086,"completion_tokens":128,"prompt_tokens_details":{"audio_tokens":0,"video_tokens":0,"cached_tokens":256,"cache_write_tokens":0},"completion_tokens_details":{"audio_tokens":0,"image_tokens":0,"reasoning_tokens":55}},"tokens_in":1086,"tokens_out":128,"cost_source":"openrouter_usage","duration_ms":2723,"image_pages":[],"temperature":1.0,"billed_attempts":1,"reasoning_tokens":55,"transport_provider":"openrouter","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-10-06T13:46:59.741896+00:00","model_set":{"reader":"deepseek/deepseek-v4-flash"},"pipeline_version":"pith-pipeline@v1.4.1-daily-deepseek-or-medium","input_kind":null,"input_chars":null,"input_total_chars":null,"passes_missing":null,"published_standing":null,"published_confidence":null,"objection_audit":null,"run_record":null,"basis":{"kind":"not_recorded","source":null,"chars_given":null,"chars_total":null},"receipt":null,"falsifier":"A decisive check: for $d = 2$, $\\mathrm{MPoly}^2 \\simeq \\mathbb{A}^1$ and $h_{\\mathrm{crit}}(a) = \\hat h_{f_a}(0)$ for $f_a(z) = z^2 + a$, so a direct test is to search over $K^{\\mathrm{ab}}$-points, for example through ramified $p$-power towers over fixed primes, for any pairwise-distinct sequence of non-PCF parameters with $\\hat h_{f_a}(0) \\to 0$, which would refute Theorem 1.1 outright; alternatively, computing the Chambert-Loir model measure on a model whose special fiber contains a component with $(H^m \\cdot Y_k) = 0$ would pinpoint whether the descent in Proposition 3.2 fails before the theorem itself does.","supporting_citations":[],"lane_accuracy":{"status":"not yet measured","lane":{"model":"deepseek/deepseek-v4-flash","input_kind":null},"url":"https://pith.science/accuracy"},"review_version":1}