{"id":"61821e5d-af31-4781-aef0-ffaa1037ea2c","arxiv_id":"2506.13517","paper_version":1,"verdict":"CONDITIONAL","confidence":"LOW","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Under the conditions K_M=O(-2D), D a smooth Calabi-Yau, and O_{3D}(D) trivial, the compact Kähler manifold M is biholomorphic to P^1 times D.","lead":"This paper proves a classification result: if a compact Kähler manifold contains a smooth Calabi-Yau divisor whose anticanonical class is exactly twice the divisor, and a certain infinitesimal triviality condition holds, then the manifold must split as a product of the Riemann sphere and the divisor. The proof builds a complete Ricci-flat metric on the complement and uses a splitting argument, extending a known program for anticanonical divisors.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The analytic decay estimates needed for the splitting are asserted rather than proved, and the weighted Sobolev inequality has an unjustified exponent range.","rationale":"The reader's weakest assumption was that the analytic machinery in Sections 3 and 4 is not fully proved, especially Proposition 3.10, Proposition 4.6, and the integral-tuning argument. I agree that the analytic core is the soft spot, but I pinpoint Proposition 4.6 as the single most load-bearing unproved step: without the weighted decay of the Monge–Ampère solution, the final Bochner-formula argument in Theorem 4.7 cannot be made rigorous. I also found an additional concrete issue not highlighted by the reader: the stated range of the weighted Sobolev inequality appears incompatible with the real dimension of the model end C×D if n denotes the complex dimension. This affects the Poisson solvability that feeds into the metric refinement. I do not see a clear counterexample or an internal contradiction that would justify rejecting the theorem outright; the argument may be repairable by choosing σ in the valid range and by supplying the missing energy estimate. Therefore the reader's CONDITIONAL verdict is retained, and no further verdict adjustment is proposed.","tokens_in":24305,"tokens_out":23518,"duration_ms":244149,"concrete_test":"Independently re-derive Proposition 4.6 for a solution u of (ω+i∂∂̄u)^n=e^fω^n: write down the energy identity on the annulus A(R,2R) and prove E(R)=∫|∇u|²=O(R^{−δ}) for some δ>0 using f∈C∞_μ and ∫(e^f−1)ω^n=0. If the nonlinear terms prevent such an estimate, compute the indicial roots of the linearized operator at infinity to see whether any positive decay rate δ is allowed; a non-positive root (for example a log term) would mean ∇u∈C∞_1 and ∇²u∈C∞_2 do not follow, so the boundary terms in Theorem 4.7 need not vanish.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The splitting theorem stands or falls on Proposition 4.6, which asserts that the solution u of the complex Monge–Ampère equation from Theorem 4.4 satisfies ∇u∈C∞_1 and ∇²u∈C∞_2, so that the Calabi–Yau metric is still ALG_{1,2}. Its proof is only a few sentences: it invokes 'the same energy estimate as in Proposition 3.12' without giving an energy identity for the nonlinear Monge–Ampère equation, and it does not prove the required C^0 decay |u−u_B|≤Cr^{−δ} with δ>0. Proposition 3.12 is a linear Poisson estimate, and the nonlinear equation does not automatically inherit that boundary identity. Without such decay, the weighted Schauder estimates cannot produce ∇u∈C∞_1 and ∇²u∈C∞_2. This is load-bearing because in Theorem 4.7 the vanishing of the boundary term ∫_{∂N_r}∇ν|∂̄v|² relies on |∂̄v|²=O(r^{−2δ}) and ∇ν|∂̄v|²=O(r^{−1−2δ}); if Proposition 4.6 fails, the conclusion ∂̄v=0 is unsupported. In addition, the weighted Sobolev inequality in Propositions 2.10 and 2.11 is stated with σ∈[1,n/(n−2)], but the end C×D has real dimension 2n when n is the complex dimension, so the Sobolev conjugate only gives σ≤n/(n−1). The rescaling argument does not improve this range. The Moser iteration in Proposition 3.10 does not explicitly check that the σ it uses is in the valid range; if it relies on the stated larger range, the Poisson solvability itself is not established. These are gaps in the proof rather than known counterexamples, but they are exactly the steps needed to obtain the ALG Calabi–Yau metric used for the splitting.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves a holomorphic splitting theorem for compact Kähler manifolds whose anticanonical divisor is twice a smooth Calabi-Yau divisor, under a triviality condition on the restriction of the normal bundle to the third infinitesimal neighborhood. The strategy is to construct a complete ALG_{1,2} Kähler metric on the complement, solve the Poisson equation to improve the decay of the Ricci potential, invoke a Tian-Yau/Hein existence theorem for the complex Monge-Ampère equation, and then use the resulting Ricci-flat metric together with a carefully chosen harmonic function to force a holomorphic splitting. The paper also gives a verification of the theorem in complex dimension two via the Enriques-Kodaira classification.","tokens_in":24674,"tokens_out":16281,"duration_ms":160992,"significance":"If the proof can be completed, the theorem gives a clean classification statement: under the stated hypotheses, the only possibility is a holomorphically trivial P^1-bundle over a Calabi-Yau base. The paper also introduces a generalized ALG framework and weighted Sobolev inequalities that could be useful for future construction of complete Calabi-Yau metrics. The argument relies on substantial external theorems (Yau, Tian-Yau, Hein, Haskins-Hein-Nordström) with independent prior publication, and I see no circularity or fitting of constants. However, the analytic core of the paper is not yet fully proved: several load-bearing estimates are asserted or only sketched. The significance of the result justifies serious revision, but the current manuscript does not meet the standard of a complete proof.","major_comments":[{"comment":"The weighted Sobolev inequality is stated for every σ∈[1,n/(n−2)] (and likewise for α in Proposition 2.11). This range is not correct: the annulus A_i=A(λ^i,λ^{i+1})×Y has real dimension 2n, since Y has real dimension 2n−2, so the standard Sobolev conjugate on this fixed 2n-dimensional domain gives only σ≤n/(n−1). The rescaling argument in the proof rescales only the Euclidean factor and cannot improve the Sobolev conjugate of the total dimension. This matters because Proposition 3.10 uses the inequality with p=2σ and needs σ<2 in the Moser iteration; for n≥3 the stated range and the valid range differ. The author should either correct the range and then verify that all choices of σ and ε used in Proposition 3.10 remain admissible, or prove a genuinely different weighted inequality.","section":"§2.3, Propositions 2.10 and 2.11"},{"comment":"The proof of Proposition 3.10 is only a sketch and contains several unverified steps. The line 'Using Prop and p = 2, λ = ū_ε' does not identify the proposition and contains the garbled term '||∇u2 2||'. The following Hölder estimate requires a proof that ||r^{−μ+1+ε}||_{(2σ)'} is finite, which imposes restrictions on σ and ε that are not stated. Later the text says 'Observed that if σ, we have' and then asserts an inequality involving ||r^{(1+ε)σ−2ε−2}||_2 without justifying the integrability condition. The claim that μ/(2σ−1) > (σ−1)/(2−σ) is automatically guaranteed by μ>2 is false for σ close to 2; for example μ=2.1 and σ=1.9 give 0.75 > 9, which is false. As written, the Moser iteration does not establish the uniform L∞ bound needed for the exhaustion argument.","section":"§3.2, Proposition 3.10"},{"comment":"Proposition 4.6 asserts ∇u∈C∞_1, ∇²u∈C∞_2 and that (M,ω_u,J) is ALG_{1,2}, but the proof is two sentences: it says the energy estimate in Proposition 3.12 applies and then lists (4.13) with δ and 1+δ instead of 1 and 2. Proposition 3.12 is a linear Poisson estimate with an explicit boundary identity; the complex Monge-Ampère equation is nonlinear, and no energy identity or C^0 decay |u−u_B|≤Cr^{−δ} is supplied for it. If δ<1, the conclusions in (4.13) are weaker than the statement and do not imply ALG_{1,2}. This gap is load-bearing because Theorem 4.7 needs the boundary integral ∫_{∂N_r} ∇ν|∂̄v|² to vanish, and the asserted decay |∂̄v|²=O(r^{−2δ}), ∇ν|∂̄v|²=O(r^{−1−2δ}) is never derived from the constructed harmonic function v and its auxiliary solution w′.","section":"§4.1, Proposition 4.6"},{"comment":"The argument that one can arrange ∫_M(e^{f_4}−1)ω_4^n=0 by adding ε i∂∂̄v_4 lacks a sign and continuity analysis. For a fixed meromorphic volume form, ∫_M(e^f−1)ω^n = ∫_M i^{n²}Ω∧Ω̄ − ∫_M ω^n, so increasing the volume form decreases the integral. The text starts from ∫(e^{f_4}−1)ω_4^n<0 and adds a positive form with i∂_{J0}∂̄_{J0}v_4=α_4>0; to leading order this makes ∫ω^n larger and therefore makes the integral more negative. In the positive case the text subtracts the same positive form, which again changes the integral in the wrong direction. Even apart from the sign, the existence of parameters (ε,r_1,r_2) giving exactly zero requires a continuity and monotonicity argument that is not provided. Since Theorem 4.4 is invoked only under the integral condition, this is a load-bearing gap.","section":"§4, integral-tuning paragraph after (4.9)"},{"comment":"Lemma 3.9 claims that solvability for f∈C∞_1 is equivalent to solvability for f′∈C∞_μ with μ>2 and ∫_M f′=0. The proof only sketches one direction on a product manifold and says 'repeat this step one more time' and 'the analysis on the new potential and some further modification follows the same routine' for the non-product case. The decomposition of f into a fiberwise average and a mean-zero part, together with the O(log r/r²) error in (3.10), is not enough to conclude that the refined potential lies in C∞_μ with μ>2; no estimate for the modified potential after the second step is given. Because Theorem 3.13 depends on this reduction, a complete argument is needed.","section":"§3.1, Lemma 3.9"}],"minor_comments":[{"comment":"There are typos in the abstract and introduction: 'volumn form' should be 'volume form', 'potiential' should be 'potential', and 'in stead' should be 'instead'.","section":"Abstract and §1"},{"comment":"The class C^{k,α}_δ is defined by C^{k,α} membership and derivative estimates, but the Hölder seminorms on the end are not explicitly controlled; since the weighted Schauder estimates later use Hölder norms, the definition should include the decay of the Hölder seminorms as well.","section":"§2.3, Definition 2.6"},{"comment":"The proof refers to 'proposition 8.11' when the intended reference appears to be Proposition 3.11 (the weighted Schauder estimate of Hein-Tosatti).","section":"§3.2, Proposition 3.12 proof"},{"comment":"In the displayed Monge-Ampère equation (ω+i∂∂̄u)^m=e^f ω^m, the exponent should be n, the complex dimension of M, rather than m, which is not defined at that point.","section":"§4, Theorem 4.4"},{"comment":"The set U_D is described as (1/v)^{-1}(B(0,1))⊂R²; the notation should be B(0,1)⊂C, since v takes values in C and the splitting is holomorphic.","section":"§4.1, proof of Theorem 4.7"}],"recommendation":"major_revision","confidential_remarks":"The theorem is interesting and the overall strategy is coherent, but the analytic core of the manuscript is not complete. I recommend major revision rather than rejection because the gaps are identifiable and likely repairable: the Sobolev exponent range can be corrected, the Moser iteration can be rewritten with explicit admissible parameters, and the Monge-Ampère decay estimates require a genuine nonlinear energy argument. The integral-tuning sign issue also needs to be fixed, but it appears to be a repairable sign or convention error. The paper's use of external theorems is independent and shows no circularity."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a plausible and potentially worthwhile β=2 analogue of HHN15, but the analytic machinery it leans on is not yet proved in the preprint. I would not desk-reject it; I would send it to a serious referee, and I would expect major revision.\n\nWhat's new: Theorem 1.2 is genuinely new. Prior work covered β=1, and here the statement is the β=2 splitting with the O_{3D}(D) triviality condition. The dimension-two sanity check via Enriques-Kodaira classification is a nice touch, and the author is honest about not knowing whether condition (3) is necessary, even giving a nontrivial extension example. The overall route follows HHN15, but that is a reasonable way to proceed, not a defect.\n\nWhat's solid: the strategy is coherent and the external tools—Tian-Yau, Hein's Monge-Ampère solvability, HHN15 gauge-fixing, Yau's theorem—are real and independently published. I see no circularity or fitted constants, and the citation pattern looks fine.\n\nWhere the soft spots are: Proposition 4.6 is the biggest problem. It asserts the decay needed for the Calabi-Yau metric to remain ALG_{1,2} by saying 'the same energy estimate as in Proposition 3.12,' but that estimate is for a linear Poisson equation, while the complex Monge-Ampère equation contains the nonlinear term Q(i∂∂̄u). Without a genuine C^0 decay estimate for u, the boundary term in Theorem 4.7 need not vanish and the splitting collapses. This is load-bearing, not cosmetic.\n\nThe stress-test concern about the weighted Sobolev exponent is also correct. If n is the complex dimension, the end C×D has real dimension 2n, so the Sobolev conjugate for an L²-gradient bound is 2n/(n−1), not 2n/(n−2). The stated range σ∈[1,n/(n−2)] in Propositions 2.10 and 2.11 is too large, and the rescaling argument does not fix it. The Moser iteration in Proposition 3.10 may only need σ near 1, so this might be repairable, but as written the proposition is false or at least unproved.\n\nProposition 3.10 is also sketched: the local energy argument is not fully written, and the passage from solutions on exhaustions to a global solution with ∫|∇u|²<∞ needs more detail. The integral-tuning step to make ∫(e^f−1)ωⁿ=0 is only outlined; the sign and continuity of A(ε,r₁,r₂) are not established. These are medium-sized gaps, not fatal contradictions. Lemma 3.9's reduction from C∞₁ to C∞_μ with mean zero is plausible but abbreviated.\n\nFor whom: researchers working on the Tian-Yau program or ALG Calabi-Yau metrics should read this, but the proof is not ready to cite. My recommendation: send to peer review with the clear expectation of major revision; the theorem deserves referee time.","headline":"A plausible new β=2 splitting theorem with a coherent strategy inherited from HHN15, but several load-bearing analytic estimates are asserted rather than proved; worth refereeing, not yet citable.","tokens_in":25220,"tokens_out":5732,"would_cite":false,"duration_ms":58539,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["53C55","32Q25","32Q20","35J96"],"pacs":[],"model":"deepseek-v4-flash","headline":"A compact Kähler manifold whose anticanonical divisor is twice a smooth Calabi-Yau divisor, with a third-order triviality condition, must be biholomorphic to P^1 times that divisor.","keywords":["holomorphic splitting theorem","Calabi-Yau manifold","anticanonical divisor","complex Monge-Ampère equation","ALG manifolds","Ricci-flat Kähler metric","Poisson equation on noncompact manifolds","P^1-bundle"],"falsifier":"A direct falsifier would be a compact Kähler manifold $M$ with a smooth divisor $D$ such that $K_M=\\mathcal{O}(-2D)$, $D$ is Calabi-Yau, and $\\mathcal{O}_{3D}(D)$ is holomorphically trivial, while $M$ is not biholomorphic to $\\mathbb{P}^1\\times D$; a concrete place to search is a nontrivial $\\mathbb{P}^1$-bundle over a Calabi-Yau base whose section has the required normal and infinitesimal triviality properties, where computing the transition functions of $\\mathcal{O}_{3D}(D)$ would settle the matter.","tokens_in":24089,"feed_emoji":"📐","tokens_out":15947,"duration_ms":134913,"temperature":0.7,"pith_summary":"The paper proves a rigidity theorem: if a compact Kähler manifold $M$ has anticanonical bundle equal to $\\mathcal{O}(-2D)$ for a smooth divisor $D$ that is itself Calabi-Yau, and if a certain third-order infinitesimal triviality condition holds, then $M$ is biholomorphic to $\\mathbb{P}^1\\times D$. The setting matters because such pairs $(M,D)$ are exactly the ones for which the complement $M\\setminus D$ carries a global nowhere-vanishing holomorphic volume form, so they are candidates for complete Ricci-flat Kähler metrics. The theorem says the only compact manifolds in this class are holomorphically trivial $\\mathbb{P}^1$-bundles over a Calabi-Yau base, so instead of constructing new examples the paper rules them out. A check in complex dimension two, using the classification of compact Kähler surfaces, confirms the pattern and shows that the one non-product candidate violates the third-order hypothesis.","feed_headline":"Double anticanonical divisor forces a Kähler manifold to split","feed_subtitle":"Complete Ricci-flat metrics exist on the complement, and nontrivial examples are ruled out.","key_machinery":"The machinery is the analytic theory of $\\mathrm{ALG}_{1,2}$ manifolds: Kähler manifolds whose end is diffeomorphic to $\\mathbb{C}\\times D$, with the metric approaching the product metric at rate $O(r^{-1})$ and the component of the complex-structure difference that controls $\\bar\\partial z$ decaying like $O(r^{-2})$, a rate supplied by triviality of $\\mathcal{O}_{3D}(D)$. On such manifolds the paper solves the Poisson equation with a three-part solution decomposition (a radial logarithmic term, a term in $C^\\infty_1$, and a bounded term with controlled first and second derivatives). That solution is used to deform the background metric so the Ricci potential $f$ lies in $C^\\infty_\\mu$ for some $\\mu>2$ with $\\int_M (e^f-1)\\omega^n=0$, which is exactly the hypothesis of the noncompact Monge-Ampère theorem that produces the Calabi-Yau metric. The final step is the Bochner-Kodaira identity for $|\\bar\\partial v|^2$ on a Ricci-flat Kähler manifold: together with the decay of $\\bar\\partial v$ at infinity it forces $\\nabla\\nabla v=0$ and $\\bar\\partial v=0$, so $v$ is a holomorphic function with parallel gradient that splits the metric, and $1/v$ compactifies the split.","core_discovery":"The central claim is Theorem 1.2: let $(M,\\omega,J)$ be a compact Kähler manifold and $D$ a smooth divisor satisfying $K_M = \\mathcal{O}(-2D)$, $D$ a compact Calabi-Yau manifold, and $\\mathcal{O}_{3D}(D)$ trivial as a holomorphic line bundle on the infinitesimal neighborhood of $D$ of order three. Then $M$ is biholomorphic to $\\mathbb{P}^1\\times D$. The proof constructs a complete Kähler metric on $M\\setminus D$ asymptotic to the product metric on $\\mathbb{C}\\times D$ with $O(r^{-1})$ decay, refines it until the Ricci potential has decay $O(r^{-\\mu})$ for some $\\mu>2$ and zero weighted average, and solves the complex Monge-Ampère equation to obtain a complete Ricci-flat Kähler metric. On this background a Bochner identity forces the Hessian of an auxiliary solution to vanish, yielding a holomorphic function $v$ with $|\\partial v|=1$ and $\\bar\\partial v=0$; the reciprocal of $v$ identifies $M\\setminus D$ with $\\mathbb{C}\\times D$, and the identification compactifies to the asserted biholomorphism. The author emphasizes the negative corollary: no nontrivial examples exist.","pith_inferences":["The author explicitly leaves open whether condition (3) is necessary; the dimension-two example suggests it is. A higher-dimensional search for manifolds satisfying (1) and (2) with $\\mathcal{O}_{3D}(D)$ nontrivial but $\\mathcal{O}_{mD}(D)$ trivial for some $m>3$ would show whether the third-order threshold is sharp.","The analytic route appears specialized to multiplicity two. For $K_M=\\mathcal{O}(-\\beta D)$ with $\\beta>2$, the decay profile of the Ricci potential and the infinitesimal condition would change, so the splitting mechanism gives no reason to expect products are forced.","Because the proof yields a parallel holomorphic vector field, the rigidity is isometric as well as holomorphic: the complete Ricci-flat metric is a cylinder $\\mathbb{C}\\times D$ with product Kähler form. That means any candidate counterexample can be tested by computing the asymptotic holonomy of its end, which would have to be trivial in the product direction."],"forward_implications":["If the theorem is correct, the only compact Kähler manifolds in this class are holomorphically trivial $\\mathbb{P}^1$-bundles over a Calabi-Yau base, so no nontrivial $\\mathbb{P}^1$-bundles or deformations occur.","For every pair satisfying the hypotheses, the complement $M\\setminus D$ admits a complete Ricci-flat Kähler metric asymptotic to the product $\\mathbb{C}\\times D$, giving a positive existence result in the multiplicity-two case of the longer construction program.","When $h^1(M)=0$, condition (3) is automatic; hence any compact Kähler $M$ with $K_M=\\mathcal{O}(-2D)$ and $D$ Calabi-Yau and $h^1(M)=0$ splits as $\\mathbb{P}^1\\times D$.","The splitting map $1/v$ is a proper holomorphic submersion from a neighborhood of $D$ onto a disk, so $D$ is a fiber of a $\\mathbb{P}^1$-fibration, not merely an abstract divisor.","In complex dimension two, the surface classification leaves only $\\mathbb{P}^1$ times a torus; the unique non-product extension has $\\mathcal{O}_{2D}(D)$ nontrivial, so it violates the third-order condition."],"supporting_citations":[{"why":"Supplies the starting method: solving the complex Monge-Ampère equation on the complement of an anticanonical divisor to obtain complete Ricci-flat Kähler metrics.","marker":"[TY90]"},{"why":"Extends the construction to the case where the divisor has multiplicity greater than one, giving the analytic template the paper adapts to multiplicity two.","marker":"[TY91]"},{"why":"Provides the gauge-fixing, holomorphic defining function, and decay estimates for the complex-structure difference that the paper uses to build the ALG background metric and the Ricci-potential estimates.","marker":"[HHN15]"},{"why":"Supplies the noncompact Monge-Ampère solvability theorem and the energy estimates used to obtain the Calabi-Yau metric on the complement.","marker":"[Hei10]"},{"why":"Provides the weighted Schauder estimate used to upgrade the derivative decay of solutions to the Poisson equation on the cylindrical end.","marker":"[HT20]"}],"fun_headline_variants":["Double anticanonical CY divisor forces P^1 split","Kähler manifold with CY twice anticanonical splits into P^1×D","No nontrivial examples: double anticanonical CY forces split","Ricci-flat metric on complement implies holomorphic splitting","Solving Monge-Ampère yields splitting: no examples survive"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof stands on the analytic claim that on the noncompact end of such a manifold the Poisson equation can be solved with the stated growth control and that the background metric can then be adjusted so the Ricci potential decays faster than $r^{-2}$ and has zero weighted average; the preprint sketches rather than fully proves these steps.","fun_headline_variants_meta":{"raw":{"variants":["Double anticanonical CY divisor forces P^1 split","Kähler manifold with CY twice anticanonical splits into P^1×D","No nontrivial examples: double anticanonical CY forces split","Ricci-flat metric on complement implies holomorphic splitting","Solving Monge-Ampère yields splitting: no examples survive"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00077,"raw_usage":{"total_tokens":3423,"prompt_tokens":967,"completion_tokens":2456,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":583,"completion_tokens_details":{"reasoning_tokens":2366}},"tokens_in":583,"tokens_out":2456,"duration_ms":16924,"temperature":1.0,"reasoning_tokens":2366,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T20:00:31.114009+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct falsifier would be a compact Kähler manifold $M$ with a smooth divisor $D$ such that $K_M=\\mathcal{O}(-2D)$, $D$ is Calabi-Yau, and $\\mathcal{O}_{3D}(D)$ is holomorphically trivial, while $M$ is not biholomorphic to $\\mathbb{P}^1\\times D$; a concrete place to search is a nontrivial $\\mathbb{P}^1$-bundle over a Calabi-Yau base whose section has the required normal and infinitesimal triviality properties, where computing the transition functions of $\\mathcal{O}_{3D}(D)$ would settle the matter.","supporting_citations":[],"review_version":2}