{"id":"790fded2-8a46-448e-a892-a1dbf8f9dd23","arxiv_id":"2506.09907","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"A compact Kähler manifold carrying a degree-2 hyperbolic cohomology class with nonzero top self-intersection must have nonzero holomorphic sections of adjoint bundles for sufficiently positive bundles, cannot be uniruled, and in dimension two must be of general type.","lead":"The paper defines a topological notion of hyperbolicity for compact Kähler manifolds, based on a cohomology class whose universal-cover lift has a bounded primitive and a nonzero top self-intersection. It shows this purely topological condition forces strong geometric consequences, including non-vanishing of holomorphic sections of adjoint bundles and negative curvature estimates.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The non-vanishing proof depends on an L2 Γ_s-index formula and a spectral passage for a U(1)-central extension on the infinite-volume universal cover; this cited step is the least secure point.","rationale":"I read Prop. 7.3 as the keystone of Theorem 7.1. Steps (1) and (2) are well supported: Theorem 5.1 supplies the spectral gap, and closed range follows from the gap. Step (3) is where external results do the heaviest lifting. The Gromov-Vafa-Witten polynomial (9) is valid only if the Γ_s-index is actually defined and computed by the local index theorem on M̃; Γ_s contains a non-discrete U(1) factor acting on the fiber, and M̃ is noncompact. The paper is transparent about relying on [Eys97, §7.2.2] and [Eys97, Prop. 7.1.2], and the introduction honestly notes the absence of examples separating topological from weak Kähler hyperbolicity. That honesty is a point in its favor, not a defect. My agreement with the reader is partial: the same step is the weakest, but the final spectral passage can likely be justified by the boundedness of η; the residual risk is the index formula itself. Since this is a citation-dependent but not internally inconsistent step, ACCEPT with moderate confidence remains appropriate; I would not change the verdict. If the cited Eys97 results fail in this context, the verdict would move to CONDITIONAL.","tokens_in":34451,"tokens_out":24859,"duration_ms":279108,"concrete_test":"Verify [Eys97, Prop. 7.1.2] and §7.2.2 against the present setup: do they permit a non-discrete central extension Γ_s and an infinite-volume base? As an independent analytic check, compute the left and right sides of (9) for M a compact complex-hyperbolic surface with E=O_M, using the explicit spectral data of the twisted spin-c Dirac operator on complex hyperbolic space, and confirm that the L2 Γ_s-index equals ∫_M Todd(M)∧ch(Λ^{m,0})∧exp(−sμ/2π); a mismatch, or a failure of the trace-class property over a fundamental domain for the U(1)-action, would invalidate Prop. 7.3(3).","verdict_should_be":"UNCHANGED","load_bearing_attack":"The keystone of Theorem 7.1 is Prop. 7.3(3): nonzero holomorphic sections are produced by twisting the spin-c Dirac operator with ∇s=∇0+i s η, where dη=π*μ, and by using the L2 Γ_s-index formula (9). Here Γ_s is a central extension 1→U(1)→Γ_s→π1(M)→1, the U(1) factor acts on the bundle fiber rather than on the base, and the base M̃ has infinite volume. The paper cites [Eys97, §7.2.2] for the heat-kernel computation of the Γ_s-index and [Eys97, Prop. 7.1.2] to pass from nonzero kernels of the twisted operators for arbitrarily small s to 0∈σ(ð_{E,m}). If those cited tools require a discrete deck group acting on a finite-volume quotient, or do not cover the fiber U(1)-action, then the polynomial index formula (9) is not established and Theorem 7.1 loses its proof. The remaining steps—closed range of the Laplacians and Nakano vanishing—are standard once the spectral input is granted. The final spectral passage is less fragile than it looks, since η∈L∞ makes D_s−D_0 a bounded perturbation; the residual risk is concentrated in the Γ_s-index computation itself.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces the notion of Kähler topologically hyperbolic manifold, defined by the existence of a degree-2 hyperbolic cohomology class with non-zero top self-intersection. It proves that this property is bimeromorphically and homotopically invariant, that such manifolds are not uniruled and are not bimeromorphic to compact Kähler manifolds with trivial first real Chern class, and that the product behaves functorially. The main analytic result is Theorem 7.1: on a Kähler topologically hyperbolic manifold, any Hermitian holomorphic vector bundle that is Nakano positive on a full-measure open subset has H^{m,0}(M, K_M ⊗ E) ≠ 0 and h^{m,0}(M, K_M ⊗ E) = χ(M, K_M ⊗ E) > 0. The proof combines a spectral gap theorem for elliptic operators on Galois coverings (Theorem 5.1), the Gromov–Vafa–Witten twist, and an L² Γ_s-index computation. The paper also derives pointwise curvature bounds for Kähler metrics on such manifolds. Claudon's appendix gives an explicit description of degree-2 hyperbolic classes for finitely presented groups and shows that Kähler topologically hyperbolic surfaces are of general type.","tokens_in":34560,"tokens_out":12582,"duration_ms":150449,"significance":"If the main theorem is correct, it is a substantial result: a purely topological condition forces non-vanishing of holomorphic sections of adjoint bundles, without any algebraic positivity assumption on the hyperbolic class itself. The definition is natural and unifies Gromov's Kähler hyperbolicity and weak Kähler hyperbolicity, and the homotopy invariance proved in Proposition 4.5 is a strong feature. The paper is largely self-contained for the spectral tool Theorem 5.1, and the appendix by Claudon provides an explicit, checkable description of hyperbolic classes in terms of ℓ∞-cocycles, which is independently useful. There are no fitted parameters in the main derivation. The weakest point is the index-theoretic step in Proposition 7.3(3), where the L² Γ_s-index formula is imported from references whose hypotheses are not stated; this is load-bearing for Theorem 7.1 and needs to be made explicit before the main claim can be considered fully established.","major_comments":[{"comment":"The L² Γ_s-index formula (9) is the keystone of Theorem 7.1, but the proof is a citation to [Eys97, §7.2.2] together with the local index theorem. The group Γ_s sits in the extension 1 → U(1) → Γ_s → π1(M) → 1, so it is not a discrete group, whereas the L²-index theory of [Ati76] invoked at [Ati76, Prop. 2.4] is normally formulated for countable discrete groups. Moreover the universal cover M̃ has infinite volume, so the heat semigroup is not trace class, and the definition of the L² Γ_s-index as an integral of heat-kernel traces over a fundamental domain requires justification. Please state the exact hypotheses of [Eys97, §7.2.2] and [Ati76, Prop. 2.4] and verify them for this non-discrete central extension acting on the line-bundle fiber.","section":"§7, Prop. 7.3(3), Eq. (9)"},{"comment":"The step 'we can find ε > 0 such that ker D^s_{E,m} ≠ {0} for each s ∈ (0, ε) and thus, thanks to [Eys97, Prop. 7.1.2], 0 ∈ σ(ð_{E,m})' is not self-contained. The cited proposition is not stated, so the reader cannot check whether its hypotheses hold when the twist is by a flat bundle with connection ∇s = ∇0 + isη on an infinite-volume cover and when the kernels are taken in the U(1)-equivariant sense. Since this is the mechanism that produces the nonzero L²-harmonic (m,0)-form, the argument should be written out or the cited result quoted with its hypotheses.","section":"§7, Prop. 7.3(3), spectral passage"},{"comment":"The assertion that Nakano positivity on a full-measure open subset A implies H^{m,q}_{∂E}(M,E) = 0 for q ≥ 1 is used to identify h^{m,0}(M,E) with the positive Euler characteristic. Standard Nakano vanishing is usually stated under positivity everywhere; by continuity the curvature is only semipositive on the zero-measure complement. Please provide a precise reference for a Demailly-type semipositive vanishing theorem or a short argument showing the vanishing under this weaker hypothesis.","section":"§7, proof of Th. 7.1"}],"minor_comments":[{"comment":"The phrase 'M is topologically Kähler hyperbolic' should read 'M is Kähler topologically hyperbolic'.","section":"§4, Prop. 4.5"},{"comment":"The final passage from the lower bound on im(E(1)) to all w ∈ D(P) is only justified by a page reference to [BDET24, p. 28]; since L does not commute with the spectral projection E(1), please expand the decomposition argument.","section":"§5, proof of Th. 5.1"},{"comment":"In the statement of Proposition A.8, the second summand should be H²_hyp(G₂,R), not H²_hyp(G,R).","section":"Appendix A, Prop. A.8"},{"comment":"The authors honestly note that no explicit examples of Kähler topologically hyperbolic manifolds beyond weakly Kähler hyperbolic ones are currently known; this limitation is worth preserving in the final version, as it contextualizes the new results.","section":"Introduction, p. 4"},{"comment":"The chain of inclusions contains a repeated equality; a short sentence indicating which inclusions are known strict and which are conjectural would improve readability.","section":"Remark 4.8"}],"recommendation":"major_revision","confidential_remarks":"For the editor: the central concern is the Γ_s-index step in Proposition 7.3(3). The paper relies on Eys97's machinery and on the same-authors preprint BCDT24 for birational invariance; if BCDT24 is not yet published, the dependency should be flagged. If the authors can supply the missing justification for the L² Γ_s-index computation and the spectral passage, I would support acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: the paper introduces Kähler topologically hyperbolic manifolds (a degree-2 real class whose pullback to the universal cover has a bounded primitive and whose top power is nonzero) and proves that this purely topological condition controls birational geometry, spectral gaps, non-vanishing of H^{m,0} for adjoint bundles, and curvature bounds. The main theorem, 7.1, is a genuine advance: the absence of positivity on the cohomology class is compensated by Nakano positivity of the bundle on a full-measure open set, and the Gromov–Vafa–Witten trick still works. The proof is a coherent chain from the spectral criterion (Thm 5.1) through closed-range arguments to the index-theoretic step.\n\nWhat's new: the definition itself, Theorem 5.1 as a general spectral gap statement for elliptic operators with zeroth-order potential, and the applications to non-uniruledness, non-Calabi–Yau bimeromorphic models, and scalar/Ricci curvature estimates. The appendix by Claudon gives a clean cochain description of degree-2 hyperbolic classes and shows surfaces in the class are of general type. The paper is well-written, organized, and unusually honest: the authors explicitly state that they have no examples separating topological hyperbolicity from weak Kähler hyperbolicity.\n\nSoft spots, in order. (1) The keystone Prop 7.3(3) is where a careful referee should focus. The Γ_s-index formula and the passage from nonzero twisted kernels to 0 ∈ spectrum are cited to [Eys97] rather than proved. This is a standard tool for the Vafa–Witten trick, and the concern that it might fail on infinite-volume universal covers is not substantiated—η is an L∞ 1-form, so the twisted operator is a bounded perturbation of the untwisted one. But because the result is cited, the paper inherits the risk. (2) Several key lemmas, including the pullback of hyperbolic classes over classifying maps, come from the same authors' BCDT24 preprint. That is not circular—[BCDT24] is independent—but a referee might want those statements stated precisely. (3) Minor typos, e.g., 'topologically Kähler hyperbolic' in Prop 4.5. Nothing load-bearing.\n\nVerdict: this deserves a serious referee. The central claim is likely correct and the ideas are transferable. I would send it out, and I expect a competent referee to come back with requested clarifications on the index step rather than a rejection.","headline":"A new 'topologically hyperbolic' class for Kähler manifolds, with strong spectral and non-vanishing consequences; the core proof uses the standard Vafa–Witten trick and the softest spot is reliance on cited index-theoretic results.","tokens_in":35244,"tokens_out":2588,"would_cite":true,"duration_ms":28081,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["32Q15","58J50","32L20","53C21"],"pacs":[],"model":"deepseek-v4-flash","headline":"A purely topological condition on Kähler manifolds forces nonzero holomorphic sections of adjoint bundles.","keywords":["Kähler topologically hyperbolic","hyperbolic cohomology class","spectral gap","Nakano positivity","Griffiths positivity","effective non-vanishing","L2 index theorem","Kodaira dimension"],"falsifier":"Compute χ(M,K_M⊗E) for a compact Kähler topologically hyperbolic manifold with a Hermitian holomorphic vector bundle Nakano positive on a full-measure open set; an example with χ(M,K_M⊗E) ≤ 0 and $H^{{m,0}}$_{∂E}(M,E)={0} would directly refute Theorem 7.1. Alternatively, exhibit a Kähler topologically hyperbolic surface of Kodaira dimension less than 2, contradicting the appendix's Theorem A.10.","tokens_in":34102,"feed_emoji":"📐","tokens_out":10798,"duration_ms":105758,"temperature":0.7,"pith_summary":"Compact Kähler manifolds are usually called hyperbolic when a Kähler class lifts to a bounded exact form on the universal cover. This paper strips the positivity away: a manifold is Kähler topologically hyperbolic when it carries any real degree-2 cohomology class whose pullback to the universal cover is d-exact with a bounded primitive, and whose top self-intersection is nonzero. The main theorem says that on such a manifold, any Hermitian holomorphic vector bundle that is Nakano positive on a full-measure open set has nonzero holomorphic sections twisted by the canonical bundle, with the dimension of those sections equal to a positive Euler characteristic. These conclusions are new even in the classical Kähler hyperbolic setting, and they lead to effective non-vanishing results for adjoint line bundles and to quantitative curvature obstructions. The appendix identifies hyperbolic degree-2 classes with those whose cocycles are bounded in the first variable for each fixed second variable, and shows that Kähler topologically hyperbolic surfaces are of general type.","feed_headline":"A topological condition forces holomorphic sections","feed_subtitle":"A bounded primitive 2-class with nonzero top power implies new sections and negative curvature.","key_machinery":"The central object is a hyperbolic cohomology class in degree 2: a class [η]∈H²(M,R) whose pullback π*η to the universal cover is exact with a bounded primitive β (dβ=π*η), together with homological non-singularity ∫_M η^m≠0. The argument is carried by the Gromov–Vafa–Witten trick: twist a spin-c Dirac operator by the family of connections ∇_s=∇_0+isη, use heat-kernel traces to compute the L²-Γ_s index, and apply the local index theorem to write that index as a polynomial in s whose leading coefficient is ∫_M η^m. Because that coefficient is nonzero, the polynomial is not identically zero, so there is an interval of s where the twisted kernels are nonzero; a comparison result then pushes zero into the spectrum of the untwisted Dirac operator. A separate spectral-gap theorem for elliptic operators on Galois coverings (Theorem 5.1) rules zero out of the spectrum on the positive-Nakano pieces, forcing the zero eigenvalue to sit in the (m,0)-part, which is exactly $H^{{m,0}}$_{∂E}(M,E).","core_discovery":"On the paper's own terms, the discovery is that the Gromov–Vafa–Witten mechanism works with only the top self-intersection of a bounded-primitive 2-class, replacing the Kähler, nef, and big positivity assumptions used in earlier notions of hyperbolicity. Theorem 7.1 states that if (M,h) is Kähler topologically hyperbolic and (E,τ)→M is a Hermitian holomorphic vector bundle that is Nakano positive over an open subset A⊂M of full measure, then $H^{{m,0}}$_{∂E}(M,E)≠{0} and $h^{{m,0}}$_{∂E}(M,E)=χ(M,K_M⊗E)>0. For line bundles this is read as a special case of the effective non-vanishing conjecture, since the Nakano-positivity hypothesis forces E to be big and nef. The same spectral control gives curvature inequalities such as min_M scal_h ≤ −4λ̃_{0,h} for weakly Kähler hyperbolic manifolds, and the appendix proves that Kähler topologically hyperbolic surfaces have Kodaira dimension 2.","pith_inferences":["Beyond the paper, if the paper's Conjecture 4.13 holds, then Kähler topological hyperbolicity would be a cohomological obstruction to being special in the sense used in the classification of Kähler manifolds.","Beyond the paper, the appendix's cochain description makes it plausible that topological hyperbolicity can be checked on the fundamental group alone for finitely presented groups, using only group-cohomology computations.","Beyond the paper, the polynomial index formula suggests a quantitative refinement: the dimension of H^{m,0}(M,K_M⊗E) may be governed by the size of ∫_M η^m and by curvature bounds, giving explicit constants in effective non-vanishing statements."],"forward_implications":["A Kähler topologically hyperbolic manifold cannot be uniruled: no dominant meromorphic map from P¹×N can cover it, and in the projective case the canonical bundle is pseudoeffective.","It cannot be bimeromorphic to a compact Kähler manifold with trivial first real Chern class, ruling out complex tori, Calabi–Yau, and hyperkähler factors up to finite étale covers.","Every Hermitian holomorphic vector bundle that is Nakano positive on a full-measure open set yields nonzero sections of K_M⊗E, and the dimension equals χ(M,K_M⊗E)>0; for line bundles, K_M⊗E^p is big and h^{m,0}(M,E^p)>0 for every positive p.","Every Kähler metric on a weakly Kähler hyperbolic manifold satisfies min_M scal_h ≤ −4λ̃_{0,h}, with equality iff scal_h is constantly −4λ̃_{0,h}; in the topologically hyperbolic case a Ricci lower bound a forces a ≤ −λ̃_{0,h}/m.","Kähler topologically hyperbolic surfaces are of general type."],"supporting_citations":[{"why":"Supplies the Kähler hyperbolicity notion and the Gromov–Vafa–Witten trick for using a bounded primitive with nonzero top self-intersection to control spectra.","marker":"[Gro91]"},{"why":"Introduces weak Kähler hyperbolicity and the spectral-gap and L² Hodge techniques that the paper generalizes to purely topological hyperbolicity.","marker":"[BDET24]"},{"why":"Provides the birational-invariance and classifying-space arguments adapted in Theorem 4.6 and Corollary 4.7.","marker":"[BCDT24]"},{"why":"Defines hyperbolic cohomology classes on general manifolds and supplies Theorem 2.4 used to transfer hyperbolic classes along classifying maps.","marker":"[BKS24a]"},{"why":"Atiyah's L²-index theorem underlies the equality h^{m,0}_{∂E}(M,E)=χ(M,K_M⊗E) via von Neumann dimensions.","marker":"[Ati76]"},{"why":"Prop. 7.1.2 is the comparison that passes from nonvanishing kernels of twisted Dirac operators to zero in the spectrum of the untwisted operator, a keystone of Proposition 7.3(3).","marker":"[Eys97]"},{"why":"Local index theorem for twisted spin-c Dirac operators used to obtain the polynomial formula (9) for the L² Γ_s-index.","marker":"[Dui11]"},{"why":"Non-amenability yielding positive bottom of the spectrum on the universal cover, used in the spectral and curvature consequences.","marker":"[Sik01]"}],"fun_headline_variants":["Kähler topological hyperbolicity forces holomorphic sections","Topological hyperbolicity yields non-vanishing sections","Spectral gap theorems from Kähler topological hyperbolicity","Bounded primitive 2-class forces sections"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof stands on the L² index-theoretic step in Proposition 7.3(3): the index of the twisted spin-c Dirac operator must coincide with the polynomial in the twisting parameter given by the local index theorem (with nonzero leading coefficient), and the cited comparison result [Eys97, Prop. 7.1.2] must force 0 into the spectrum of the untwisted operator on the infinite-volume universal cover.","fun_headline_variants_meta":{"raw":{"variants":["Kähler topological hyperbolicity forces holomorphic sections","Topological hyperbolicity yields non-vanishing sections","Spectral gap theorems from Kähler topological hyperbolicity","Bounded primitive 2-class forces sections"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00088,"raw_usage":{"total_tokens":3820,"prompt_tokens":976,"completion_tokens":2844,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":592,"completion_tokens_details":{"reasoning_tokens":2779}},"tokens_in":592,"tokens_out":2844,"duration_ms":28713,"temperature":1.0,"reasoning_tokens":2779,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T04:39:24.130891+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute χ(M,K_M⊗E) for a compact Kähler topologically hyperbolic manifold with a Hermitian holomorphic vector bundle Nakano positive on a full-measure open set; an example with χ(M,K_M⊗E) ≤ 0 and $H^{{m,0}}$_{∂E}(M,E)={0} would directly refute Theorem 7.1. Alternatively, exhibit a Kähler topologically hyperbolic surface of Kodaira dimension less than 2, contradicting the appendix's Theorem A.10.","supporting_citations":[],"review_version":1}