{"id":"8827b626-887a-4290-9df0-de6db13b1b64","arxiv_id":"2607.14342","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Connected labeled Eulerian maps with arbitrary even degree distribution scale as V! V^{(5g-7)/2} e^{VΩ}, with constants depending only on mean degree and mean squared degree.","lead":"This paper finds a universal large-size asymptotic for labeled Eulerian maps of any genus and any mix of vertex degrees, plus an exact count at genus one. The leading constants are governed by the Painlevé I recurrence, connecting map enumeration to integrable systems.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The ACSV saddle-point step for mixed valences is not fully justified: nonnegativity of the relevant A_g coefficients and strict minimality of t(α) are asserted, not proved; a failure would change the leading constant.","rationale":"The reader's verdict is CONDITIONAL and identifies the same weak point. I agree this is the most load-bearing concern. The paper's overall strategy is coherent: the Riemann-Hilbert analysis gives explicit r_g, the structural lemma (Prop 5.21) gives the square-root singularity, and the ACSV machinery is standard. There is independent support: the planar formula (1.19) is recovered, and regular map cases reduce to known results, so the theorem is not circular. However, the novel mixed-valence step rests on the ACSV hypotheses. The nonnegativity of the coefficients in (5.25) is explicitly flagged in the paper as 'the most restrictive' and is disposed of by assertion; the minimality of t(α) is proved in one sentence. Neither is a trivial detail: ACSV minimal-point theory requires strict minimality, and the constants K_g are determined by the local singularity at the unique minimal point. A second minimal point or nonpositive coefficients would change the constant factor, though the exponent and e^{VΩ} might survive. I would not move to REJECT because the formula is corroborated in regular/large classes and the gap is plausibly fillable; but the proof as written is conditional on these unproved analytic-combinatorial facts. The proposed p=2 test is a concrete way to check whether the concern is real. If the test failed, the central theorem would be in jeopardy; if it passed, the conditional verdict could be upgraded. Secondary issues (the 0/0 in Theorem 2 and the g=0 case for Δ=5g-1=-1) are also present, but they do not threaten the core g≥1 asymptotic as directly as the ACSV step.","tokens_in":58380,"tokens_out":23386,"duration_ms":215799,"concrete_test":"Specialize to p=2, α=(1/2,1/2). Compute the discriminant ξ(t_2,t_4)=Res_σ(P,Pσ) explicitly (a polynomial in t_2,t_4). Enumerate all solutions of ξ=0 with |t_2|≤|t_2(α)| and |t_4|≤|t_4(α)| using, e.g., resultant isolation or numerical algebraic geometry. If any solution other than t(α) satisfies both coordinate-wise bounds, t(α) is not strictly minimal and the ACSV step fails. Independently, extract the first 50 coefficients of A_1(t) in the basis of (5.25) from (2.35) and (5.92), and check they are nonnegative. This directly tests the two unproved hypotheses in the simplest mixed-valence case.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The proof of Theorem 1 is carried by Proposition 5.29, whose hypotheses include the non-negativity condition (5.25) for the coefficients of Q/Pσ^Δ and strict minimality of the point t(α) on the singular variety. The paper asserts after (5.25) that non-negativity 'comes for free' from combinatorial interpretations (r_g(t) counts 2-legged maps), but no bijection or sign-cancellation proof is supplied for the specific functions A_g(t) that enter via Proposition 5.33. The positivity of [t^n]A_g = E(E+1)[t^n]F_g follows from Proposition 5.32 only if one already knows that F_g has positive coefficients in this basis; this is true by (1.24), but the paper does not spell out the chain for A_g. More importantly, Proposition 5.28 gives only a one-sentence extremum argument for the minimality/uniqueness of t(α). The ACSV deformation in Proposition 5.29 is valid only if t(α) is strictly minimal; if another singularity lies inside the coordinate-wise torus, the contour can be pulled to S_loc only up to an unquantified contribution, and the leading constant K_g would be altered. Since this step is used for every g≥1, it is the most load-bearing assumption in the paper.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies connected, labeled Eulerian maps of fixed genus g with arbitrary mixed even-degree sequences. Theorem 1 states a universal asymptotic as the total number of vertices V tends to infinity: the leading term is K_g/Γ((5g-1)/2) · V^{(5g-7)/2} · V! · e^{VΩ(α)}, with K_g depending on the degree proportions α only through ε and ζ and satisfying a Painlevé-I-type recurrence. Theorem 2 gives an exact formula for genus 1 Eulerian maps. The proof combines the Hermitian matrix-model topological expansion with orthogonal-polynomial recurrence coefficients, the discrete string equation, and analytic combinatorics in several variables (ACSV). The paper also recovers Tutte's planar formula and several known regular-valence higher-genus results.","tokens_in":58668,"tokens_out":15361,"duration_ms":153569,"significance":"If the main theorem is correct, it is a substantial advance: it provides the first higher-genus asymptotic enumeration for mixed-valence Eulerian maps, identifies a universal dependence on only two linear functionals of the degree distribution, and connects the constants to Painlevé I. The exact genus-1 formula and the recovery of known regular cases are concrete, useful checks. The proof strategy is appropriate and much of the derivation is explicit, with substantial technical appendices. The main concerns are not with the overall architecture but with three load-bearing points: the final coefficient-extraction factor, the proof of uniqueness/minimality of the ACSV critical point, and the nonnegativity hypothesis for the singular generating functions.","major_comments":[{"comment":"As printed, the transition from Proposition 5.32 to Eq. (5.94) has a factor error. Proposition 5.32 gives [t^n]F_g = [t^n]A_g / (E(E+1)), and E ~ εV, so the displayed chain immediately before (5.94) should have [t^{αV}]F_g = [t^{αV}]A_g / (ε²V²), not multiplied by ε²V². Correspondingly, (5.94) as written — N_g = V! I_V(...) ε²V² — produces V-exponent (5g+1)/2 rather than the theorem's (5g−7)/2. If the intended formula is V! I_V/(ε²V²), all occurrences should be corrected; as it stands, the printed proof is inconsistent with the statement of Theorem 1.","section":"§5.3, proof of Theorem 1, Eq. (5.94)"},{"comment":"The proof of the existence and uniqueness of the minimal point t(α) is too sketchy for a step that carries the entire ACSV argument. The one-sentence extremum calculation only checks that t(α) is a critical point of the height function restricted to the parametrized component Σ_cr. It does not rule out other components of the discriminant variety (such as the higher-degeneracy points mentioned in Remark 5.11) lying inside the coordinate-wise torus, nor does it prove global uniqueness of the minimum of the height function. Strict minimality is used in Proposition 5.29 to discard the contribution from the outer contour S_0 as exponentially small; without it the leading constant K_g could be altered. Please supply a complete proof of uniqueness/minimality, or state precisely which theorem from [52]/[53] applies and verify its hypotheses.","section":"§5.2.2, Proposition 5.28"},{"comment":"The nonnegativity condition f(n)≥0 is asserted to 'come for free' from combinatorial interpretations, but it is not established for the specific functions A_g(t)/P_σ^{5g−1} that feed into Theorem 1 through Proposition 5.33. The chain A_g = Q(Q+1)F_g together with the combinatorial definition of F_g would imply f(n)=E(E+1)N_g(n) once Proposition 5.32 is applied, but this is not written out; for r_g(t) the paper relies on a cited 2-legged-map interpretation. Since both Proposition 5.27 and the contour deformation in Proposition 5.29 depend on this positivity in the chosen basis, the paper should give an explicit coefficient-positivity proof, or a precise reference that covers exactly these algebraic functions.","section":"§5.2, Eq. (5.25)"}],"minor_comments":[{"comment":"The notation I_V( ε^g(t(α))^{5g−1} / (κ(ε−1)) C_g; ...) is ambiguous: the symbol g appears both as the genus index and as a potential function g(t) introduced in Corollary 5.18. Please disambiguate, e.g. denote the genus by g and the function by G(t).","section":"§5.3, Eq. (5.94)"},{"comment":"The displayed formula has several unmatched large brackets and the multiplication by the sum is difficult to parse. Re-typesetting the bracketed expression and indicating the scope of the product would substantially improve readability.","section":"Theorem 2, display"},{"comment":"The sentence 'positive orthant in their proposition has been replaced by negative orthant here' is informal. Since the signs in the generating function are absorbed by the variables s_{2k} = -t_{2k}/(2k), state explicitly how [52, Prop. 3.17] is applied after this change of variables.","section":"§5.2, Proposition 5.27"},{"comment":"There are numerous small OCR/typo artifacts in the equations (e.g. missing backslashes, odd superscripts such as 'ε^g(t(α))'). A careful proofreading pass is needed before publication.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The paper is promising and the main theorem is likely correct, but I cannot recommend acceptance in the present form. The final factor error, even if it is partly a typesetting artifact, must be corrected in all displayed equations; the ACSV minimality and nonnegativity hypotheses are load-bearing and need complete proofs rather than one-sentence assertions. None of these issues seems impossible to fix within the scope of the manuscript, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is the first genuinely mixed-valence Eulerian map result for genus ≥ 1, and it deserves a careful referee. The main asymptotic (Theorem 1) is plausible and largely well-supported; the exact genus-1 formula has a well-posedness bug as displayed, but the alternative Proposition 4.9 looks right.\n\nWhat is actually new: prior explicit work covered regular Eulerian maps, and here arbitrary degree proportions α, fixed genus g ≥ 1, lead to a universal leading term depending only on ε and ζ, with constants governed by a Painlevé-I recurrence. The route is legitimate: topological expansion from [24], Riemann–Hilbert/orthogonal-polynomial analysis, the string equation to extract singular structure of the recurrence coefficients, then ACSV. They do real work, recover Tutte’s planar formula and known regular cases as sanity checks, and the derivations are explicit enough to check.\n\nSoft spots, in proportion:\n\n1. Theorem 2 as displayed has a genuine 0/0 problem. The sum over 0 < r < n includes terms with E(r) = 1, and if the factor E(r)(E(r)-1) is in the denominator, those terms are undefined. Proposition 4.9 gives a cancellation-free formula that avoids the issue, but the paper asserts equivalence without proving it, and the stated theorem needs a caveat or a repaired formula. This is the clearest defect, and it is fixable.\n\n2. The ACSV step rests on two assertions that are true but under-explained. Nonnegativity of the coefficients in (5.25) for the specific functions A_g(t) is not immediate from the combinatorial remark given; it does follow from Proposition 5.32 plus positivity of N_g, but the chain is not spelled out. Strict minimality of t(α) is dispatched with a one-sentence extremum argument. I suspect it is correct, but the leading constant depends on it, so a referee should ask for a real proof.\n\n3. Theorem 1 states g ≥ 0, but the ACSV machinery as formulated applies to A_g with denominator P_σ^{5g-1}, which for g = 0 has no pole. The g = 0 case needs a separate argument, or the theorem should say g ≥ 1 and relegate g = 0 to the known Tutte asymptotic. Minor, because genus 0 is already solved, but the statement overclaims as written.\n\n4. The theorem assumes all α_{2k} > 0; regular and mixed cases with zero proportions are handled only by informal limiting statements. Also minor, but worth a remark.\n\nFor the right audience — enumerative combinatorists and random matrix theorists — this is a substantial contribution. I would send it to peer review, expecting revision rather than rejection, and I would ask the authors to fix the genus-1 formula and tighten the ACSV justification.","headline":"First mixed-valence Eulerian map asymptotics for genus ≥ 1, credible and worth refereeing, but the displayed genus-1 formula has a small well-posedness bug and a couple of proof details need tightening.","tokens_in":59157,"tokens_out":7578,"would_cite":true,"duration_ms":77281,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["05C30","05A15","05A16","60B20","42C05"],"pacs":[],"model":"deepseek-v4-flash","headline":"One generating-function singularity governs the asymptotic count of all mixed-valence Eulerian maps, for every fixed genus.","keywords":["Eulerian maps","map enumeration","mixed valence","asymptotic enumeration","universality","random matrix models","orthogonal polynomials","Painlevé I"],"falsifier":"Take a degree mixture with vertices of degrees 4 and 6 only, compute the first few coefficients of the genus-one generating function A_1(t), and check that none are negative and that the discriminant surface has no second minimal point with the same coordinate-wise absolute values as t(α); a negative coefficient or a competing minimal point would break the saddle-point derivation.","tokens_in":58231,"feed_emoji":"🗺️","tokens_out":5393,"duration_ms":58610,"temperature":0.7,"pith_summary":"This paper aims to settle the leading-order enumeration of connected, labeled Eulerian maps whose vertices may have many different even degrees. It claims that once the proportions of vertices of each degree are fixed, the count grows as a universal power of V times V! times an exponential factor, with a prefactor that depends on the degree mixture only through two numbers: the limiting average degree and the limiting sum of squared degrees. The genus-dependent prefactor constants obey the same nonlinear recurrence that produces Painlevé I coefficients. If this is right, it closes a previously open problem for mixed-valence maps of positive genus and also delivers an exact closed-form count for genus one.","feed_headline":"One formula counts all even-degree maps on every surface","feed_subtitle":"Leading term depends only on two degree statistics; the genus constants follow a Painlevé-I recurrence.","key_machinery":"The central object is the algebraic curve P(σ;t)=κ that defines the equilibrium-measure endpoint σ(t) for a Hermitian matrix model with even potential. The recurrence coefficients r_g(t) of the associated orthogonal polynomials are shown to be rational in σ and derivatives of P, and near the critical surface they have the singular form r_g(t) = C_g ξ(t)^{-(5g-1)/2}(a_g(t)+b_g(t)ξ(t)^{1/2}), where ξ is the discriminant of P as a polynomial in σ. The critical surface is parameterized by the vertex-degree proportions α, with t(α) the unique minimal point; near it, σ behaves like a(t)+b(t)ξ^{1/2}, and the Hessian determinant of the exponential tilt has an explicit closed form. This discriminant","core_discovery":"Theorem 1 asserts that for fixed genus g and fixed positive proportions α of vertices of even degrees 2,4,...,2p, the number N_g(⌊αV⌋) of connected, labeled Eulerian maps is K_g / Γ((5g-1)/2) · V^{(5g-7)/2} · V! · e^{VΩ(α)} (1+O(V^{-1/2})). Here K_g depends on α only through ε and ζ, the limiting densities of the edge count and the Zagreb index, and these constants satisfy the Painlevé-I recurrence displayed in (1.5)-(1.6). The algebraic exponent (5g-7)/2 is universal across the entire family. The same machinery reproduces Tutte's planar formula and yields Theorem 2, a new exact formula for N_1(n) for any degree sequence; the paper also notes that the same constant pattern appears in 3-regul","pith_inferences":["If the paper's Conjecture 1 for odd-degree maps holds, the leading asymptotic law for all maps would be governed by the same pair of statistics and the same Painlevé-I constants, making map enumeration asymptotically one-parameter in essence.","The structural grading introduced in Section 5.1 suggests that the subleading corrections in V^{-1/2} are themselves organized by the same recurrence, so a full asymptotic expansion may be derivable from the same discriminant geometry.","The positivity and minimality gap could be checked numerically for small mixed-degree vectors; if it fails for some α, a more careful treatment of multiple critical points would be needed, possibly changing the prefactor but not the exponent.","The exponential-rate formula Ω(α) resembles a Shannon-type entropy with a binomial correction, hinting at a large-deviation principle for map degree sequences; if developed, the asymptotic count could yield probability distributions of vertex degrees in random large maps."],"forward_implications":["All mixed-valence Eulerian degree-proportion vectors with the same ε and ζ share the same leading constant K_g; the exponential rate Ω(α) is the only mixture-sensitive part.","Regular even-valence formulas for arbitrary 2p-regular maps, and the quartic and sextic cases, reduce to Theorem 1 by setting ζ=ε².","The exact genus-one formula in Theorem 2 gives a closed count for any degree sequence with at least one vertex of degree greater than 2; for a single valence it reduces to earlier regular-map formulas.","The ratio t_g defined in Remark 1.4 is independent of ε and ζ, so the Painlevé-I constants in this paper coincide, up to known factors, with constants appearing in earlier map-asymptotic work."],"fun_headline_variants":["Universal law counts even-degree maps on every genus","Even-degree maps: one formula for all genera","Painlevé I appears in universal map-counting constants","Exact genus-1 count for Eulerian maps of any degree","Two statistics determine all genus Eulerian map asymptotics"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The argument assumes that the power-series coefficients used in the final saddle-point step are all nonnegative, so that the single dominant singularity really controls the count; the paper states this follows from a combinatorial interpretation but does not prove it for the specific series involved.","fun_headline_variants_meta":{"raw":{"variants":["Universal law counts even-degree maps on every genus","Even-degree maps: one formula for all genera","Painlevé I appears in universal map-counting constants","Exact genus-1 count for Eulerian maps of any degree","Two statistics determine all genus Eulerian map asymptotics"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001022,"raw_usage":{"total_tokens":4131,"prompt_tokens":711,"completion_tokens":3420,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":455,"completion_tokens_details":{"reasoning_tokens":3352}},"tokens_in":455,"tokens_out":3420,"duration_ms":26422,"temperature":1.0,"reasoning_tokens":3352,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T02:21:34.239042+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a degree mixture with vertices of degrees 4 and 6 only, compute the first few coefficients of the genus-one generating function A_1(t), and check that none are negative and that the discriminant surface has no second minimal point with the same coordinate-wise absolute values as t(α); a negative coefficient or a competing minimal point would break the saddle-point derivation.","supporting_citations":[],"review_version":1}