{"id":"3d1b5656-a961-4596-a3b2-0c91e181d730","arxiv_id":"2506.00456","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For every PCF polynomial, the arboreal Galois image is contained in a newly defined sign-kernel overgroup of the tree automorphism group; for a new cubic example the image is exactly E^2_n(3).","lead":"This paper studies the Galois groups that arise from repeatedly pulling back a point under a post-critically finite polynomial, and shows these groups always sit inside certain tree automorphism groups defined by sign restrictions. It also computes the rank and normal subgroup structure of one such group and gives a new polynomial whose Galois image is exactly one of these groups.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1.1's proof asserts α nonperiodic gives a full d-ary tree; this is false when α lies in the postcritical set, so the theorem needs the standard squarefree/avoid-postcritical hypothesis.","rationale":"The reader's weakest-assumption analysis identifies the same load-bearing gap: Theorem 1.1 only assumes α is not periodic, but the arboreal embedding needs f^n(z) − α squarefree for all n, i.e., α outside the postcritical set. This is exactly where the proof's first sentence breaks. The concern is substantive because it affects precisely the PCF maps with strictly preperiodic critical points that the F^{(m,m')} overgroups are designed to cover. The fix (adding α ∉ P_f) is standard and likely compatible with the intended argument, so the correct verdict remains CONDITIONAL rather than REJECT. Other issues—the off-by-two exponent in the proof of Theorem 1.1 ('m = L + 2O + 1' instead of 'L + 2O − 1'), mislabelled lemma references, and Theorem 1.3 relying on a deferred argument from [2]—are secondary and repairable. I therefore do not move the reader's verdict, and I agree that the missing hypothesis is the weakest load-bearing point.","tokens_in":21756,"tokens_out":8993,"duration_ms":92198,"concrete_test":"Set K = Q, f(z) = z^2 − 2, α = −2. Compute f(z) − α = z^2 and f^n(z) − α for n = 1, 2; observe repeated roots at every level. Verify that Gal(f(z) − α/Q) is trivial while Aut(T_1(2)) has order 2, so the claimed containment in Theorem 1.1 is not well-defined for this α. If the authors amend Theorem 1.1 to require α ∉ P_f, re-run the proof of Lemma 2.9 and the induction step under that hypothesis; the counterexample is then excluded and the discriminant-based sign argument is non-vacuous.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The proof of Theorem 1.1 opens: 'Since α is not periodic, the dynamical tree of f at α is isomorphic to a full d-ary tree.' That is false. A full d-ary tree requires f^n(z) − α to have d^n distinct roots for every n, i.e., α must avoid the forward orbit of every finite critical point. The theorem only excludes periodic α. A concrete failure: f(z) = z^2 − 2 over Q has C_f = {0}, postcritical orbit 0 → −2 → 2 → 2, and α = −2 is strictly preperiodic, hence not periodic. But f(z) − α = z^2 has a double root, so the level-1 preimage tree is not a binary tree and Gal^1_f(α) does not embed into Aut(T_1(2)) in the usual sense. Lemma 2.9's discriminant formula then contains the factor (f(c) − α) = 0, so 'disc is a square' is vacuous and imposes no sign condition on a well-defined full-tree action. Since PCF maps with strictly preperiodic critical points have finite postcritical sets, this is not an edge case. The theorem must add the hypothesis that α ∉ P_f (or at least α ∉ f^n(C_f) for all n), as the Introduction's 'under certain conditions' hints but Theorem 1.1 omits.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces two families of subgroups E and F of Aut(T_n(d)), defined as kernels of sign functions on the tree automorphism group, and claims (Theorem 1.1) that the n-th arboreal Galois group of any postcritically finite polynomial over a number field embeds into one of these groups, with parameters determined by the tail length L and period O of the critical set. The paper further studies the group E^2_n(d), proving it has rank 2 and a unique chief series (Theorem 1.2), and presents f(z)=2z^3-3z^2+1 as an example whose arboreal image is exactly E^2_n(3) under certain arithmetic conditions (Theorem 1.3). The manuscript is written from the author's thesis and leans heavily on earlier work, especially [2], for both framework and proof details.","tokens_in":22070,"tokens_out":14426,"duration_ms":137283,"significance":"If the main theorem were fully established, it would be a valuable structural result: all PCF arboreal representations would be constrained by finitely many quadratic sign conditions, and the exact realization E^2_n(3) would be a concrete new example of an arboreal image equal to one of the overgroups. The parameter-free construction of the sign-kernel groups is a strength, and the group-theoretic analysis of E^2_n is potentially of independent interest. However, the current manuscript has several load-bearing gaps: Theorem 1.1 omits the standard postcritical-avoidance hypothesis, its proof contains an internal parameter inconsistency and an omitted even-degree case, Theorem 1.3 is essentially delegated to [2], and Lemma 3.6 appears false. These issues prevent acceptance in the present form, though most appear fixable.","major_comments":[{"comment":"Theorem 1.1 as stated omits the standard hypothesis that alpha avoids the postcritical set of f. The proof begins \"Since alpha is not periodic, the dynamical tree of f at alpha is isomorphic to a full d-ary tree,\" which is false: a full d-ary tree requires f^n(z)-alpha to be squarefree for every n, equivalently alpha not in f^n(C_f) for any n. For example, f(z)=z^2-2 over Q has C_f={0}, f(0)=-2, f(-2)=2=f(2), and alpha=-2 is strictly preperiodic but not periodic; however f(z)-alpha=z^2 has a double root, so Gal^1_f(alpha) does not embed in Aut(T_1(2)) in the usual sense. Moreover Lemma 2.9's discriminant formula contains the factor (f(c)-alpha), so the square-discriminant conclusion becomes vacuous when alpha is a critical value. The theorem must add the hypothesis that alpha is not in the postcritical set P_f (or at least alpha is not in f^n(C_f) for any n), as the Introduction's vague \"under certain conditions\" hints but does not state. This is load-bearing because the induction and the containment are defined only on a full d-ary tree.","section":"Theorem 1.1; §3.1 proof"},{"comment":"The statement and proof of Theorem 1.1 do not match in two places. For odd L>1, the proof chooses m=L+2O+1 and cites Lemma 2.9.(2), but the theorem states F^{(L+2O-1,L-1)}_n(d), and the applicable Lemma 2.9.(3) gives exponent L+2O-1. In the same paragraph the conclusion is phrased as containment in E^{(m,m')}_m, although the theorem requires F for this case. For even degree, the theorem only lists L=0 and L>1, omitting L=1; Lemma 2.9.(4) covers L=1 and would give (m,m')=(O+1,1). This is not an empty case: for f(z)=z^4-2z^2, C_f={0,±1}, f(C_f)={0,-1} is periodic, but C_f itself is not, so L=1. The theorem needs a complete case split, and the proof must use the matching parameters.","section":"Theorem 1.1(2)-(3); §3.1 proof"},{"comment":"The proof of Theorem 1.3 consists of the sentence \"In [2], the authors showed that the existence of elements of order 2 and 3 implies the presence of a two-cycle permutation in Gal(K^2_f/K) and a three-cycle permutation in Gal(K^3_f/K) that fixes f^{-1}(x). The same proof works in our case.\" This is not a proof of the exact isomorphism Gal^n_f(alpha) ≅ E^2_n(3). The preceding lemmas establish irreducibility and the existence of elements of order 2 and 3 in certain extensions, and Theorem 1.1 gives containment in E^2_n(3), but equality requires an order computation or an explicit argument that the group generated has order |E^2_n(3)|. The appeal to [2] does not verify that its hypotheses transfer to f(z)=2z^3-3z^2+1 and to Condition (3.1). Please provide the missing details, or state Theorem 1.3 as conditional on the argument in [2] applying verbatim.","section":"Theorem 1.3; §3.3 proof"},{"comment":"Lemma 3.6 claims H_i=H_1 for all i with k_i=d^{i-1}+1. This is not proved and appears false. The group Aut(T_n) preserves the level-1 block partition of the leaves, so the pair {1,k_1} lies inside a single level-1 block, while {1,k_2} with k_2=d+1 lies in two different level-1 blocks; these pairs are in different orbits. For n=2 and d=3, the subgroup generated by X_1 is contained in the tuples that have even parity in each level-1 block, whereas X_2 contains the cross-block vector e_1+e_4, which is not in that subspace; hence H_1≠H_2. Since Lemma 3.8 invokes Lemma 3.6 to identify ker(res_{n-1})/M_n as the unique minimal normal subgroup of E^2_n/M_n, the proof of Theorem 1.2 depends on this step and needs to be repaired or replaced.","section":"Lemma 3.6; §3.2"}],"minor_comments":[{"comment":"The statement says \"with m>n,\" but the proof uses Aut(T_{n-m}), so the intended hypothesis is m≤n.","section":"Lemma 2.1"},{"comment":"In the induction step, \"the last equality follows from n being even\" should read \"d being even.\"","section":"Proposition 2.3 proof"},{"comment":"The proof says \"The proof follows by induction, and more details can be found in [2].\" Please provide the induction or a precise statement of which result in [2] is being transferred and why it applies here.","section":"Corollary 3.11"},{"comment":"There are numerous typos and formatting issues, including \"proof Theorem 1.1\" in the section heading, \"a f n\" in the abstract, and the acknowledgments thanking \"the anonymous referee ... and its eventual publication,\" which is inappropriate for a submitted manuscript. A careful editorial pass is needed.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"To the editor: The manuscript appears to be an early-stage write-up. The proof of Theorem 1.3 is largely quoted from [2], and the false Lemma 3.6 suggests that the group-theoretic part needs a careful rewrite. The missing postcritical-avoidance hypothesis in Theorem 1.1 is a standard and fixable condition, and the parameter inconsistency is also fixable. If the author can supply the missing hypotheses, correct the parameters, and repair or replace Lemma 3.6, the paper could be a solid contribution. In its current form I cannot recommend publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Wayne Peng's paper proposes a uniform constraint on arboreal Galois images of postcritically finite polynomials: every image sits inside one of the sign-kernel groups E or F, with parameters coming from the critical orbit tail length and period. That is a genuinely new statement, extending the Belyi-map results in [2] and [3] to all PCF polynomials. The two-parameter sign kernels are a natural construction, and the rank-2/unique-chief-series result for E^2_n (Theorem 1.2) is a real group-theoretic contribution. The example f(z)=2z^3-3z^2+1, with critical portrait 0<->1, is new and worth knowing.\n\nThe main problem is that Theorem 1.1 is false as stated. The proof opens by asserting that alpha nonperiodic implies the preimage tree is a full d-ary tree. That is not true: the tree is full only if f^n(z)-alpha is squarefree for every n, which requires alpha to avoid the forward orbit of every finite critical point. Example: f(z)=z^2-2, alpha=-2 is strictly preperiodic but not periodic, and f(z)-alpha = z^2 has a double root. The introduction's 'under certain conditions' hints at this, but the theorem omits it. This is a load-bearing omission, though easily repaired by adding alpha not in P_f.\n\nThe proof also has internal mismatches. For odd degree and L>1, the proof cites Lemma 2.9.(2), which is the even-degree, all-critical-points-periodic case; the correct reference is Lemma 2.9.(3). And it sets m=L+2O+1 where the discriminant statement needs L+2O-1. These are fixable typos, but they make the current version untrustworthy.\n\nTheorem 1.3 is presented as a new example, but the proof is a one-paragraph appeal to [2] rather than a self-contained argument. The attached lemmas give the Eisenstein and ramification part, but the final step, that order-2 and order-3 elements force the full E^2_n image, is deferred.\n\nThe citation pattern is fine; the author leans on [2] and his own thesis only where the results are specifically relevant. The paper is not circular.\n\nBottom line: the core idea is plausible and potentially useful, but the current manuscript is not publishable as is. If the author adds the standard postcritical-avoidance hypothesis, corrects the lemma references and the exponent, and fleshes out Theorem 1.3, it deserves a serious referee, and I would then be happy to see it in print.","headline":"Plausible overgroup theorem for PCF arboreal images, but Theorem 1.1 needs a postcritical-avoidance hypothesis and several proof fixes.","tokens_in":22602,"tokens_out":3721,"would_cite":false,"duration_ms":32810,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["11R32","37P05","11R58"],"pacs":[],"model":"deepseek-v4-flash","headline":"Every PCF polynomial's Galois tree image sits in a small sign-kernel overgroup, and one new polynomial hits it exactly.","keywords":["arboreal Galois representation","postcritically finite polynomial","PCF polynomial","d-ary tree","sign-kernel overgroup","wreath product","chief series","dynamical Belyi polynomial"],"falsifier":"For a specific degree-3 PCF polynomial whose critical set has tail length L = 0 and period O = 1, compute the index-two sign-kernel group $E^{2}$_n(3) and the order of Gal^n_f($\\alpha$) for one small n; if Gal^n_f($\\alpha$) contains an element of odd sign at level 2 while also failing to lie in ker(sgn_2), the containment in $E^{2}$_n(3) fails. More directly, choose $\\alpha$ equal to a critical value of f, so f(z) - $\\alpha$ has a repeated root; then $Gal^{1}$_f($\\alpha$) is not even well-defined as a subgroup of Aut(T_1(d)), and the claimed embedding, as stated, breaks.","tokens_in":21565,"feed_emoji":"🌳","tokens_out":2010,"duration_ms":21753,"temperature":0.7,"pith_summary":"This paper tries to show that the arboreal Galois image of any postcritically finite (PCF) polynomial is never the full automorphism group of the d-ary tree; instead it must lie inside one of finitely many explicitly defined sign-kernel overgroups determined only by the tail length and period of the critical set. The author then isolates a concrete cubic polynomial whose entire arboreal image is exactly one of these overgroups, giving the first sharp realization result for a PCF polynomial with the critical portrait 0 maps to 1, 1 maps to 0. A sympathetic reader would care because this turns a recurring observation in arithmetic dynamics, that PCF images have infinite index, into a precise structural statement about where the image must sit, and because it provides a group-theoretic template for computing dynamical Galois groups.","feed_headline":"PCF polynomial tree images trapped in tiny sign-kernel groups","feed_subtitle":"New theorem pins every postcritically finite arboreal Galois image inside a finite list of sign kernels, and one cubic hits its bound…","key_machinery":"The machinery is the sign-kernel overgroups of the d-ary tree automorphism group: $E^{{(m,m')}}$_n(d) and $F^{{(m,m')}}$_n(d) are defined as kernels of sign functions $sgn^{{(m,m')}}$_1 and $sgn^{{(m,m')}}$_2, which combine the natural sign on level m with the sign on level m'. For PCF polynomials, the refined discriminant formula of Proposition 2.7 shows that disc(f^n(z)-$\\alpha$) is a square in the ground field (or a fixed small extension) once n passes a bound depending only on the critical set's tail and period; this produces the sign conditions that force the Galois image into the overgroup. A second mechanism is the chief-series analysis of $E^{2}$_n(d), where the layers M_n, ker(res_n), and quotients isomorphic to C_2 wreath Aut(T_{n-1}) give a unique tower of normal subgroups and the rank bound two.","core_discovery":"The central claim is Theorem 1.1: if f is a degree d PCF polynomial over a number field K and $\\alpha$ is not periodic under f, then Gal^n_f($\\alpha$) is isomorphic to a subgroup of one of the sign-kernel groups $E^{{(m,m')}}$_n(d) or $F^{{(m,m')}}$_n(d), where m and m' are computed from the critical set's tail length L and period O. Odd-degree maps with L <= 1 land in $E^{{2O}}$_n(d); odd-degree maps with L > 1 land in $F^{{(L+2O-1, L-1)}}$_n(d); even-degree maps land in $E^{{(m,m')}}$_n(d) with (m,m') = (O+1,1) when L=0 and (L+O, L) when L>1. The proof mechanism is that for a PCF map, the discriminants disc(f^n(z) - $\\alpha$) become squares after a bounded number of levels, so quadratic sign conditions along the tree hold from some level onward and force the image into the kernel of a suitable sign homomorphism. The paper also proves that the group $E^{2}$_n(d) is generated by two elements and has a unique chief series for odd d, and exhibits f(z) = $2z^{3}$ - $3z^{2}$ + 1 as the first proven exact realization: under a mild arithmetic condition on the basepoint, Gal^n_f($\\alpha$) is isomorphic to $E^{2}$_n(3) for every n.","pith_inferences":["The paper's Theorem 1.1 likely holds under the additional hypothesis that alpha avoids the postcritical set of f, since the embedding of Gal^n_f(alpha) into Aut(T_n(d)) requires f^n(z) - alpha to be squarefree at every level; the stated 'alpha not periodic' condition alone does not guarantee this.","The exact realization for the portrait 0 -> 1, 1 -> 0 suggests that other non-conjugate critical portraits, such as 0 -> 1, 1 -> 0 with different ramification, could produce exact realizations of E^{(m,m')}_n or F^{(m,m')}_n for other small parameters, providing a testable family of examples.","The chief-series technique for E^2_n may generalize to E^m_n only at the cost of losing uniqueness of the minimal normal subgroup; the paper's own final section identifies why m >= 3 breaks the tower argument, so the rank question for those groups is open and the methods here do not transfer directly.","The semiconjugate observation that f = -B(z) + 1 with f^2 = B^2 suggests that arboreal images may be compared through functional equations, hinting at a dynamical isogeny principle for Galois images of twisted Belyi maps that the paper leaves for future work."],"forward_implications":["The Odoni index of every PCF polynomial arboreal representation is infinite, now as a corollary of containment in sign-kernel overgroups whose orders are strictly smaller than the full tree automorphism group.","For PCF polynomials, the infinite tree image is constrained by finitely many quadratic conditions, so one can hope to compute the exact image once those quadratic extensions and a finite base level are understood.","The exact realization Gal^n_f(alpha) = E^2_n(3) for f(z) = 2z^3 - 3z^2 + 1 shows that at least one natural overgroup is not merely an upper bound but the true image, validating the overgroup framework as a classification tool.","Since E^2_n(d) needs only two generators for every n and has a unique chief series, the profinite limit of the realized image is a finitely generated pro-2-type group with a rigid normal-subgroup structure.","For odd-degree normalized dynamical Belyi polynomials, the image is automatically a subgroup of E^{2}_n(d), giving a uniform structural result across that whole family."],"supporting_citations":[{"why":"Provides the definition and properties of the natural sign function on tree automorphisms and the framework for arboreal Galois representations that the overgroups E^m_n and F^{(m,m')}_n build on.","marker":"[15]"},{"why":"Supplies the large arboreal representation result for the cubic PCF polynomial and the Eisenstein/ramification arguments that Theorem 1.3's proof adapts.","marker":"[2]"},{"why":"Gives the dynamical Belyi map methods, including geometric Galois group computations, that the author proposes for answering whether Gal^n_{-B+1}(alpha) equals E^2_n or E^{(2,1)}_n.","marker":"[3]"},{"why":"Supplies the rank lemma used to bound d(E^2_n) once a unique minimal normal subgroup is identified in the chief-series argument.","marker":"[7]"},{"why":"Provides the definition and background on chief series of finite groups that Theorem 1.2's proof uses.","marker":"[13]"},{"why":"Provides prior discriminant iteration formulas over which Proposition 2.7 refines the sign computation for PCF polynomials.","marker":"[6]"},{"why":"Establishes the Odoni index and prime-density motivation that the PCF infinite-index corollary and the overgroup viewpoint address.","marker":"[22]"},{"why":"Provides the prior analysis of quadratic PCF arboreal images that motivates asking which portraits in degree 3 give exact sign-kernel images.","marker":"[24]"}],"fun_headline_variants":["PCF arboreal images confined to sign-kernel overgroups","Galois images of PCF polynomials lie in sign kernels","Exact realization: cubic hits sign-kernel bound","Tiny sign kernels house all PCF tree Galois images"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The theorem assumes that the basepoint $\\alpha$ is not periodic, but the containment statement relies on f^n(z) - $\\alpha$ being squarefree for every n, namely that $\\alpha$ avoids the forward orbits of all critical points; without that the tree collapses and the embedding into Aut(T_n(d)) is not defined.","fun_headline_variants_meta":{"raw":{"variants":["PCF arboreal images confined to sign-kernel overgroups","Galois images of PCF polynomials lie in sign kernels","Exact realization: cubic hits sign-kernel bound","Tiny sign kernels house all PCF tree Galois images"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000231,"raw_usage":{"total_tokens":1560,"prompt_tokens":1091,"completion_tokens":469,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":707,"completion_tokens_details":{"reasoning_tokens":401}},"tokens_in":707,"tokens_out":469,"duration_ms":4778,"temperature":1.0,"reasoning_tokens":401,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T12:08:02.536910+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For a specific degree-3 PCF polynomial whose critical set has tail length L = 0 and period O = 1, compute the index-two sign-kernel group $E^{2}$_n(3) and the order of Gal^n_f($\\alpha$) for one small n; if Gal^n_f($\\alpha$) contains an element of odd sign at level 2 while also failing to lie in ker(sgn_2), the containment in $E^{2}$_n(3) fails. More directly, choose $\\alpha$ equal to a critical value of f, so f(z) - $\\alpha$ has a repeated root; then $Gal^{1}$_f($\\alpha$) is not even well-defined as a subgroup of Aut(T_1(d)), and the claimed embedding, as stated, breaks.","supporting_citations":[{"cited_title":"Théorie des Nombres et Applications","cited_arxiv_id":null,"evidence_quote":"Provides the definition and properties of the natural sign function on tree automorphisms and the framework for arboreal Galois representations that the overgroups E^m_n and F^{(m,m')}_n build on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the dynamical Belyi map methods, including geometric Galois group computations, that the author proposes for answering whether Gal^n_{-B+1}(alpha) equals E^2_n or E^{(2,1)}_n."},{"cited_title":"Finite groups that need more generators than any proper quotient","cited_arxiv_id":null,"evidence_quote":"Supplies the rank lemma used to bound d(E^2_n) once a unique minimal normal subgroup is identified in the chief-series argument."},{"cited_title":"Martin Isaacs","cited_arxiv_id":null,"evidence_quote":"Provides the definition and background on chief series of finite groups that Theorem 1.2's proof uses."},{"cited_title":"Ramification in iterated towers for rational functions","cited_arxiv_id":null,"evidence_quote":"Provides prior discriminant iteration formulas over which Proposition 2.7 refines the sign computation for PCF polynomials."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the Odoni index and prime-density motivation that the PCF infinite-index corollary and the overgroup viewpoint address."},{"cited_title":"Profinite iterated monodromy groups arising from quadratic morphisms with infinite postcritical orbits, 2013","cited_arxiv_id":null,"evidence_quote":"Provides the prior analysis of quadratic PCF arboreal images that motivates asking which portraits in degree 3 give exact sign-kernel images."}],"review_version":1}