{"id":"a23ccae3-4dec-421d-a641-2fbb98aab80b","arxiv_id":"1908.03955","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A Poisson-Kähler fibration has a canonical Kähler metric on its base whose holomorphic bisectional curvature is non-positive and whose holomorphic sectional, Ricci, and scalar curvatures are bounded above by a negative constant.","lead":"This paper studies families of Kähler manifolds and proves that a special class, called Poisson-Kähler fibrations, always admit a Weil-Petersson type metric with strong negative curvature on the base. It also characterizes when a projective bundle fibration is of this type, linking the condition to projectively flat vector bundles.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem C's converse is false as stated: Higgs-flatness of A reduces to integrability of the horizontal distribution, which does not imply the Monge-Ampère equation; a trivial product fibration gives a counterexample.","rationale":"The Reader's weakest_assumption points to Theorem 7.1, the unpublished Chern-connection/adjoint input for the curvature proof of Theorem A. That is a legitimate concern about reliance on unpublished work, but it is not a demonstrated falsehood. In contrast, the proof of Theorem C in §7.3 contains a concrete logical gap that the Reader also noticed: [V_j,\\bar V_k]=0 implies only dω'=0 by Proposition 2.2(4), not the Monge-Ampère condition c_j\\bar k=0. This gap is load-bearing because Theorem C is advertised in the abstract as the existence criterion for Monge-Ampère fibrations. The product example shows the converse is actually false, not merely unproved. The central curvature theorem Theorem A is not directly invalidated: the forward direction Poisson–Kähler ⇒ [V_j,\\bar V_k]=0 used in §4 is correct, and §7.4 gives an independent curvature computation. I therefore do not recommend REJECT; the paper should be accepted conditionally on correcting or softening Theorem C, and on supplying the missing Theorem 7.1 verification or an independent proof of the adjoint property. This is why I mark agreement_with_reader as partial: the Reader's rationale flagged the Theorem C gap, but their formally identified weakest assumption was the unpublished Theorem 7.1 input for Theorem A.","tokens_in":40850,"tokens_out":19227,"duration_ms":200277,"concrete_test":"Verify the product counterexample explicitly. Take B a compact Kähler manifold of positive dimension, F an n-dimensional compact Kähler manifold, X=B×F, choose a Kähler form α on B and set ω_B=2α, ω_X=p^*α+ω_F, and ω=ω_X-p^*ω_B=ω_F-p^*α. Compute the horizontal lifts V_j with respect to ω: because there are no mixed dt∧d\\barζ terms, V_j=∂/∂t_j, so κ_j=0 and the Higgs field θ vanishes. The Lie-derivative connection ∇=D+θ+\\barθ is flat since the product is trivial, so A is Higgs-flat. But ω^{n+1}=(n+1)(-p^*α)∧ω_F^n≠0, so p is not Poisson–Kähler. This directly contradicts the 'only if' direction of Theorem C. If the computation is reproduced, Theorem C must be repaired by replacing 'Poisson–Kähler' with dω'=0 or by adding c_j\\bar k=0 as an extra hypothesis.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The load-bearing flaw is in Theorem C, proved in §7.3. The converse direction only derives [V_j,\\bar V_k]=0 from Higgs-flatness of A. By Proposition 2.2(4), this is equivalent to dω'=0, i.e. integrability of the horizontal distribution, not to the defining Monge-Ampère condition (ω_X-p^*ω_B)^{n+1}=0, equivalently c_j\\bar k≡0 (Definitions 1.1 and 2.6). The proof in §7.3 explicitly concludes Poisson–Kähler from [V_j,\\bar V_k]=0 by invoking Proposition 2.2, but that proposition gives only dω'=0; c_j\\bar k may be a non-zero basic (1,1)-form on B. The gap is not merely technical: let X=B×F, with B a positive-dimensional compact Kähler manifold, F an n-dimensional compact Kähler manifold, and choose Kähler forms α on B and ω_F on F. Set ω_X=p^*α+ω_F and let ω_B be a Kähler form on B different from α, say ω_B=2α. Then the relative form ω=ω_X-p^*ω_B=ω_F-p^*α has a nonzero horizontal component, so ω^{n+1}=(n+1)(-p^*α)∧ω_F^n≠0; p is not Poisson–Kähler. But the product has no mixed dt∧d\\barζ terms, so the horizontal lifts are V_j=∂/∂t_j, giving κ_j=∂V_j=0, θ=0, and ∇ is just the flat Lie-derivative connection of the trivial product bundle. Hence A is Higgs-flat in the paper's own sense (Proposition 7.6). Thus Theorem C's 'only if' direction is false as stated, and the abstract's equivalence 'relative Kähler fibration is Monge-Ampère iff the associated Higgs bundle is Higgs-flat' is also false.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies relative Kähler fibrations and two Weil–Petersson type metrics on the base. Its main curvature theorem (Theorem A / Theorem 4.1) states that for a Poisson–Kähler fibration with injective Kodaira–Spencer map, the non-harmonic Weil–Petersson metric is Kähler, has holomorphic sectional curvature bounded above by -2/(n|X_t|), and has non-positive holomorphic bisectional curvature. The paper also proves a characterization of projectively flat vector bundles in terms of Poisson–Kähler projective bundle fibrations (Theorem B / Theorem 6.1) and claims an equivalence between a relative Kähler fibration being Monge–Ampère and an associated infinite-rank Higgs bundle being Higgs-flat (Theorem C). The arguments include a finite-dimensional Higgs-bundle treatment of Burns' theorem, three proof strategies for Theorem A, several examples, and an appendix containing the proof of Theorem C.","tokens_in":41232,"tokens_out":13205,"duration_ms":135679,"significance":"If Theorem A can be established rigorously, it gives a positive answer to the negative-curvature problem for every Poisson–Kähler fibration, with an explicit and uniform negative upper bound for holomorphic sectional curvature. The third proof of Theorem A in §7.4 is a concrete Schumacher-type computation that, if correct, would be an independent route to this result. Theorem B is a clean and attractive characterization of projectively flat Hermitian vector bundles through Poisson–Kähler projectivizations. However, the paper's advertised equivalence Theorem C is false as stated, and the Higgs-bundle proof of Theorem A in §4.1 contains an identity that is inconsistent with the contraction action of the Kodaira–Spencer operators. These issues affect the abstract's central existence claim and the presentation of the main proof, so the manuscript cannot be accepted in its current form.","major_comments":[{"comment":"The 'only if' direction of Theorem C is false as stated. The proof derives [V_j,\\bar V_k]≡0 from Higgs-flatness and then invokes Proposition 2.2 to conclude that ω is Poisson–Kähler. Proposition 2.2(4), however, only equates [V_j,\\bar V_k]≡0 with dω'=0, where ω'=ω-c(ω); it does not imply ω^{n+1}=0 unless the geodesic curvature form c(ω) vanishes. A concrete counterexample is X=B×F with B and F compact Kähler, ω_X=p^*α+ω_F, and ω_B=2α. Then ω=ω_X-p^*ω_B=ω_F-p^*α is a relative Kähler form with c(ω)=-α≠0, so p is not Poisson–Kähler since ω^{n+1}=(n+1)(-p^*α)∧ω_F^n≠0. But the horizontal lifts commute, giving κ_j=0, θ=0, and a flat Lie-derivative connection on the product bundle, so A is Higgs-flat in the paper's own sense. The equivalence stated in the abstract and in Theorem C must be corrected, for example by adding the condition c_{j\\bar k}≡0 or by replacing 'Poisson–Kähler' with integrability of the horizontal distribution.","section":"§7.3 and abstract (Theorem C)"},{"comment":"The Higgs-bundle proof of Theorem A uses the identity -[Θ_{j\\bar k},κ_l]=κ_jκ_kκ_l+κ_lκ_kκ_j after observing that κ_jκ_l=0 on A^1. With the contraction action of κ_j defined in §7.2.1, both κ_jκ_k and κ_lκ_kκ_j vanish on A^1, so the displayed identity is inconsistent with the definition of the Kodaira–Spencer operators as contraction operators. Consequently the lower bound (4.5) does not follow from (4.3)–(4.4). The third proof in §7.4 treats κ_j as a TX_t-valued endomorphism and uses a different multiplication; the two frameworks must be reconciled before the §4.1 proof can be regarded as valid.","section":"§4.1, Eqs. (4.3)-(4.5)"},{"comment":"The proof of Theorem A, in both the §4.1 Higgs-bundle computation and the computations in §7.4 that invoke Theorem 7.2, depends critically on Theorem 7.1, which asserts that the Lie-derivative connection D is the Chern connection on each A^{p,q} and that κ_j=κ_j^*. This theorem is not proved in the present paper; it is cited to [49] and to an early version of [10]. Because the negativity of the curvature of ω_DF in the Higgs-bundle argument rests on the adjoint identity κ_j=κ_j^*, the manuscript should either provide a proof of Theorem 7.1 or point to a stable, citable version of the result.","section":"§7.1.2, Theorem 7.1"}],"minor_comments":[{"comment":"The paper uses two closely related but not identical definitions of Poisson–Kähler: (ω_X-p^*ω_B)^{n+1}=0 for fibrations between Kähler manifolds, and ω^{n+1}=0 for a relative Kähler form. This is understandable, but the two should be explicitly reconciled in a remark, since the distinction is relevant to the error in Theorem C.","section":"Definition 1.1 and Definition 2.6"},{"comment":"The sentence 'ω' is Poisson–Kähler if and only if the horizontal distribution associated to ω is integrable' is confusing, since ω' is always vertical in the sense that (ω')^{n+1}=0 by Proposition 2.2(2). The remark should be rephrased to avoid suggesting that integrability of the horizontal distribution is equivalent to the Monge–Ampère equation.","section":"§2.2, Remark 1 after Definition 2.6"},{"comment":"The inequality ∂²⟨κ_j,κ_j⟩ ≥ 2||κ_jκ_j||² appears to require the first term in (4.4) to be nonnegative; this is precisely the point at issue in the second major comment. Even if the sign is ultimately correct, the derivation needs to be rewritten so that the domain of the operators and the definition of the product are unambiguous.","section":"§4.1, Eq. (4.5)"},{"comment":"The abstract contains a typo ('homogenous'), and the spelling 'Monge-Ampère' is inconsistent with the body's 'Monge–Ampère'. These should be corrected in revision.","section":"Abstract and typography"}],"recommendation":"major_revision","confidential_remarks":"The falsehood of Theorem C is a serious issue: it is a central advertised result and the proof in §7.3 makes an explicit invalid inference from Proposition 2.2(4). The editor may wish to verify whether the third proof of Theorem A in §7.4 is fully independent of the flawed identities in §4.1; if it is, the curvature theorem may be salvageable, but the paper's abstract and existence claims need substantial correction."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, the stress-test note is right, and it is not a technical nitpick. In Section 7.3 the authors derive [V_j,\\bar V_k]=0 from Higgs-flatness and then invoke Proposition 2.2 to conclude Poisson–Kähler. But Proposition 2.2(4) only says [V_j,\\bar V_k]=0 iff d\\omega'=0, where \\omega'=\\omega-c(\\omega). That is integrability of the horizontal distribution; it makes \\omega' a Poisson–Kähler form, not \\omega. The actual condition is c_{j\\bar k}\\equiv 0. The product counterexample is clean: take X=B\\times F with \\omega_X=p^*\\alpha+\\omega_F and \\omega_B=2\\alpha. The relative form \\omega_F-p^*\\alpha fails the Monge–Ampère equation, but the horizontal lifts have no mixed terms, so all Kodaira–Spencer operators vanish and the associated bundle A is Higgs-flat. The advertised equivalence in the abstract is therefore false.\n\nThat said, the paper is not a throwaway. Theorem 4.15's general second variation formula for the non-harmonic Weil–Petersson metric is a real, non-circular computation, and the negativity estimates for Poisson–Kähler fibrations are supported by multiple independent routes, including a direct Schumacher-style calculation. Theorem B—projective flatness of E iff P(E)->B admits a Poisson–Kähler structure—is a substantive new characterization, and the proof via direct image curvature is a serious argument. These are the parts worth keeping.\n\nThe biggest problem beyond Theorem C is that Theorem 7.1 (the Lie-derivative connection is Chern and \\kappa_j=\\kappa_j^*) is imported from an unpublished preprint and an early version of [10]. That adjoint property is the technical heart of the curvature negativity, and a referee cannot verify it from this paper without access to the source. Several other references are also \"preprint\" or \"TBA,\" which makes a full check harder.\n\nMy overall read: Theorem A and Theorem B have a good chance of surviving, but Theorem C cannot stand as written. The paper deserves a serious referee rather than a desk reject, but only after the abstract and Theorem C are corrected or removed and the unpublished inputs are made available. This is a conditional major revision, not a reject outright.","headline":"Theorem A's curvature result looks real and Theorem B is a nice characterization, but Theorem C's converse is false as stated and the paper leans on unpublished foundational input.","tokens_in":41767,"tokens_out":4330,"would_cite":false,"duration_ms":48748,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["32G20","53C55","53D20"],"pacs":[],"model":"deepseek-v4-flash","headline":"For a Poisson–Kähler fibration, the base carries a Kähler metric whose holomorphic sectional curvature is bounded above by a negative constant and whose holomorphic bisectional curvature is non-positive.","keywords":["Monge-Ampère fibration","Poisson-Kähler fibration","non-harmonic Weil-Petersson metric","negative curvature","Higgs bundle","Kodaira-Spencer map","horizontal lift","relative Kähler fibration"],"falsifier":"Compute the non-harmonic Weil–Petersson curvature directly for a non-isotrivial Poisson–Kähler family of elliptic curves and compare it with the predicted upper bound $-2/n\\,|X_t|^{-1}$; any local direction with positive holomorphic bisectional curvature, or with sectional curvature above that bound, would disprove the theorem. A second check is to test the identity $\\kappa_j=\\kappa_j^*$ on the infinite-rank bundle $A^{p,q}$ for such a family, since that identity is the proof's load-bearing input.","tokens_in":40617,"feed_emoji":"📐","tokens_out":16639,"duration_ms":154143,"temperature":0.7,"pith_summary":"This paper answers a form of the negative curvature problem in Kähler geometry for a special class of fibrations. The class, called Poisson–Kähler fibrations, consists of relative Kähler fibrations whose total-space form satisfies the homogeneous Monge–Ampère equation, $\\omega^{n+1}\\equiv 0$. The paper proves that the base of such a fibration, whenever the Kodaira–Spencer map is injective, carries a canonical Kähler metric—the non-harmonic Weil–Petersson metric—whose holomorphic sectional curvature is bounded above by $-2/n\\,|X_t|^{-1}$ and whose holomorphic bisectional curvature is non-positive. It also characterizes Poisson–Kähler fibrations through projectively flat vector bundles and through flatness of an associated infinite-rank Higgs bundle.","feed_headline":"Monge–Ampère fibrations get negative-curvature bases","feed_subtitle":"A canonical Kähler metric makes the base have nonpositive bisectional and strongly negative sectional curvature.","key_machinery":"The argument is carried by an infinite-rank quasi-vector bundle $A^{p,q}$ whose fiber over $t\\in B$ is the space of smooth $(p,q)$-forms on the fiber $X_t$, equipped with a Lie-derivative connection $\\nabla$ built from the horizontal lifts of vector fields on the base. The decisive structural input, quoted from earlier work rather than proved here, is that the induced connection $D$ is the Chern connection with respect to the fiberwise $L^2$ metric and that the Kodaira–Spencer operators satisfy $\\kappa_j=\\kappa_j^*$. These identities turn the Chern-curvature identities for $D$ into pointwise estimates such as $|\\kappa_j\\kappa_j|^2\\ge |\\kappa_j|^4/(n|X_t|)$, which give the negative upper bounds on the holomorphic sectional curvature and the non-positivity of the bisectional curvature. A companion formula expresses the non-harmonic Weil–Petersson metric through the relative canonical bundle and the fiberwise scalar curvature.","core_discovery":"The central claim is Theorem 4.1: for every Poisson–Kähler fibration $p\\colon (X,\\omega)\\to B$ with injective Kodaira–Spencer map, the non-harmonic Weil–Petersson metric $\\omega_{DF}$ is Kähler, its holomorphic sectional curvature is at most $-2/n\\,|X_t|^{-1}$, and its holomorphic bisectional curvature is non-positive; here $n$ is the fiber dimension and $|X_t|$ the fiber volume. The same infinite-rank Higgs-bundle machinery yields two structural results. A holomorphic vector bundle over a compact Kähler manifold admits a projectively flat Hermitian metric exactly when its projectivized bundle is Poisson–Kähler, and over a compact curve this condition is equivalent to polystability. And a relative Kähler fibration is Poisson–Kähler exactly when the associated infinite-rank Higgs bundle, whose fibers are smooth differential forms on the fibers of the fibration, is Higgs-flat.","pith_inferences":["If the imported metric-connection and self-adjointness identities hold in wider generality, Theorem 4.15's explicit curvature formula suggests a route to negative curvature for arbitrary relative Kähler fibrations in which the geodesic-curvature terms $c_{j\\bar k}$ are controlled by a degenerate Monge–Ampère-type equation.","The Higgs-flat characterization recasts the homogeneous Monge–Ampère equation as a flatness condition on an infinite-rank Higgs bundle, suggesting that existence of Poisson–Kähler structures may be governed by a Hermitian–Einstein-type stability condition.","The explicit examples—families of elliptic curves, Kähler metric geodesics, convex function geodesics, and Hermitian form geodesics—give concrete test cases where the $-2/n\\,|X_t|^{-1}$ bound could be checked numerically or analytically.","For higher-dimensional bases, the projective-bundle characterization suggests testing whether Poisson–Kähler is equivalent to a slope-stability condition for vector bundles, extending the curve case."],"forward_implications":["Every Poisson–Kähler fibration with injective Kodaira–Spencer map has a base metric satisfying the negative curvature property, so the negative curvature problem is settled affirmatively on this entire class.","The non-harmonic Weil–Petersson metric is a genuinely Kähler metric, not merely a positive definite form, making it a usable canonical metric on bases of such fibrations.","A holomorphic vector bundle over a compact Kähler base is projectively flat (Hermitian) if and only if its projective bundle fibration is Poisson–Kähler; over a curve this is the same as polystability.","Poisson–Kähler fibrations are exactly those whose associated infinite-rank Higgs bundle is Higgs-flat, giving a flatness criterion for the homogeneous Monge–Ampère equation on fibrations.","In the one-dimensional fiber case with positive relative canonical bundle, the general curvature formula reproduces the classical negative curvature bound for the moduli space of curves."],"supporting_citations":[{"why":"It supplies the Monge–Ampère foliation notion and the finite-dimensional negative-curvature phenomenon that motivates the generalization.","marker":"[13]"},{"why":"It gives an independent proof of the main negative-curvature theorem and an equivalent definition of the class of fibrations.","marker":"[8]"},{"why":"It proves the metric-connection and self-adjointness statement for the Lie-derivative connection used throughout the curvature computations.","marker":"[49]"},{"why":"It provides an early version of the same metric-connection/self-adjointness input and a generalized curvature formula for the relative canonical bundle.","marker":"[10]"},{"why":"It supplies the flat Higgs-bundle curvature argument and Hodge-metric method that the infinite-rank proof follows.","marker":"[50]"},{"why":"It gives the curvature formula for the relative canonical bundle and computational lemmas used in the third proof of the main theorem.","marker":"[39]"},{"why":"It supplies a generalized relative-canonical-bundle curvature formula used in the curvature identity.","marker":"[33]"},{"why":"It provides the direct-image curvature theorem used to show that Poisson–Kähler projective bundles give projectively flat vector bundles.","marker":"[7]"}],"fun_headline_variants":["Monge–Ampère base curvature turns negative","Negative curvature forced on Monge–Ampère bases","Monge–Ampère fibrations yield negatively curved bases","Higgs-flatness characterizes Monge–Ampère fibrations","Monge–Ampère fibrations: negative curvature and Higgs-flatness"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole curvature estimate rests on an imported theorem, not proved in this paper, that the Lie-derivative connection on each infinite-rank form bundle is the Chern connection and that the Kodaira–Spencer operators satisfy $\\kappa_j=\\kappa_j^*$; if that fails, the negativity argument collapses.","fun_headline_variants_meta":{"raw":{"variants":["Monge–Ampère base curvature turns negative","Negative curvature forced on Monge–Ampère bases","Monge–Ampère fibrations yield negatively curved bases","Higgs-flatness characterizes Monge–Ampère fibrations","Monge–Ampère fibrations: negative curvature and Higgs-flatness"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001078,"raw_usage":{"total_tokens":4517,"prompt_tokens":958,"completion_tokens":3559,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":574,"completion_tokens_details":{"reasoning_tokens":3470}},"tokens_in":574,"tokens_out":3559,"duration_ms":27222,"temperature":1.0,"reasoning_tokens":3470,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:58:11.542227+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the non-harmonic Weil–Petersson curvature directly for a non-isotrivial Poisson–Kähler family of elliptic curves and compare it with the predicted upper bound $-2/n\\,|X_t|^{-1}$; any local direction with positive holomorphic bisectional curvature, or with sectional curvature above that bound, would disprove the theorem. A second check is to test the identity $\\kappa_j=\\kappa_j^*$ on the infinite-rank bundle $A^{p,q}$ for such a family, since that identity is the proof's load-bearing input.","supporting_citations":[{"cited_title":"Burns, Curvatures of Monge–Ampere F oliations and Parabolic Manif olds, Ann","cited_arxiv_id":null,"evidence_quote":"It supplies the Monge–Ampère foliation notion and the finite-dimensional negative-curvature phenomenon that motivates the generalization."},{"cited_title":"Berndtsson, Negative curvature and complex structures, preprint","cited_arxiv_id":null,"evidence_quote":"It gives an independent proof of the main negative-curvature theorem and an equivalent definition of the class of fibrations."},{"cited_title":"Notes on variation of Lefschetz star operator and $T$-Hodge theory","cited_arxiv_id":"1708.07332","evidence_quote":"It proves the metric-connection and self-adjointness statement for the Lie-derivative connection used throughout the curvature computations."},{"cited_title":"Curvature restrictions on a manifold with a flat Higgs bundle","cited_arxiv_id":"1608.00777","evidence_quote":"It supplies the flat Higgs-bundle curvature argument and Hodge-metric method that the infinite-rank proof follows."},{"cited_title":"Schumacher, Positivity of relative canonical bundles and applications , Invent","cited_arxiv_id":null,"evidence_quote":"It gives the curvature formula for the relative canonical bundle and computational lemmas used in the third proof of the main theorem."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It supplies a generalized relative-canonical-bundle curvature formula used in the curvature identity."},{"cited_title":"Berndtsson, Strict and non strict positivity of direct image bundles , Math","cited_arxiv_id":null,"evidence_quote":"It provides the direct-image curvature theorem used to show that Poisson–Kähler projective bundles give projectively flat vector bundles."}],"review_version":1}