{"id":"1c29dad3-465f-4239-afd9-b398b7b6d413","arxiv_id":"2608.01519","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Big line bundles on compact complex spaces satisfy a Steenbrink-type cohomology vanishing theorem after passing to a log resolution and twisting by multiplier ideal sheaves.","lead":"This paper extends Steenbrink's vanishing theorem, a classical zero-cohomology result for ample line bundles, to big line bundles on compact complex spaces using multiplier ideal sheaves and resolutions of singularities. A correct proof would give a useful general tool for showing cohomology groups vanish in the same bidegree range as the Kodaira-Nakano theorem.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proof of Theorem 1.3 applies [HLWY23, Thm 1.1] to a Q-line bundle with fractional snc twist; not shown valid, so compact-case vanishing isn't established as written.","rationale":"The paper proposes a natural generalization, and Theorem 4.1 gives a detailed L2 proof for singular positive line bundles on weakly pseudoconvex relatively compact spaces. That proof constructs an explicit complete metric and uses Hörmander's estimates, so the approach is credible. However, the compact-case proof of Theorem 1.3 is shorter and depends on applying [HLWY23, Theorem 1.1] to a Q-line bundle Lδ whose positivity includes a fractional snc twist O(Σ(1-δ_j)E_j). This is precisely the kind of step that needs justification: the cited theorem is stated for genuine line bundles (as evidenced by its use in the introduction for ample A), and it is not automatic that the positivity of Lδ transfers to vanishing for the untwisted integral sheaf. The reader's verdict of CONDITIONAL is appropriate. I do not see a separate, more serious flaw; the L2 route is a plausible repair, and the numerous self-citations to unpublished preprints, while a hindrance to verification, are not by themselves a mathematical objection. The set-theoretic typos in the descent construction (e.g., writing V\\Exc(µ) when Exc(µ) lies in V̂) are presentation issues, not central. Hence the identified fractional-twist step is the single load-bearing concern.","tokens_in":17348,"tokens_out":14664,"duration_ms":155928,"concrete_test":"Verify the exact statement of [HLWY23, Theorem 1.1]: does it apply to positive Q-line bundles of the form O(-E)⊗M⊗O(Σ(1-δ_j)E_j), where 0<1-δ_j<1? If not, check whether the proof of Theorem 4.1, run with Y=X̂ (smooth), X=X̂, V=X̂, L=µ*L, h=Ĥ, produces a complete Poincaré-type Kähler metric on X̃\\E whose associated L2 Dolbeault resolution yields exactly H^q(X̃, Ω^p(logE)⊗O(-E)⊗π̃*L⊗I(π̃*Ĥ))=0. If the L2 argument succeeds, the central claim survives and only the short proof needs repair; if it fails at the fractional twist, Theorem 1.3 as stated is unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In the proof of Theorem 1.3 (p.14), the author defines Lδ = π̃*L ⊗ I(π̃*h) ⊗ O_X̃(Σ_{j∈J}(1-δ_j)E_j) = O_X̃(-E) ⊗ π̃*L ⊗ I(π̃*h) ⊗ O_X̃(Σ_{j∈J}(1-δ_j)E_j), with 1-δ_j ∈ (0,1] because δ_j ∈ [0,1). The proof then says 'applying [HLWY23, Theorem 1.1] to the positivity of Lδ' yields the desired vanishing H^q(X̃, Ω^p(logE) ⊗ O(-E) ⊗ π̃*L ⊗ I(π̃*h)) = 0. The gap: [HLWY23, Theorem 1.1], as used elsewhere in the paper (Introduction, consequence for ample A), is a vanishing theorem for holomorphic line bundles. Lδ is a Q-line bundle, not an integral line bundle, because the twist O(Σ(1-δ_j)E_j) is generally fractional. The cited theorem's hypotheses are not shown to accommodate such a twist, and clearing denominators does not directly yield vanishing for the untwisted integral sheaf. No auxiliary argument (e.g., the L2 Poincaré-type metric construction of Theorem 4.1) is supplied in this proof. Thus the main compact-case theorem is not proven by the written reasoning.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper aims to extend Steenbrink's vanishing theorem for ample line bundles on projective varieties to big line bundles on compact complex spaces, with multiplier ideal sheaves. The main Theorem 1.3 states: if X is a compact complex space of pure dimension n, μ: \\hat X → X is a resolution, and L → X is a big holomorphic line bundle, then X is Moishezon and there is a singular Hermitian metric h on L such that μ*h has algebraic singularities off Exc(μ) and strictly positive curvature; for a log resolution π: \\tilde X → \\hat X of the singular locus of μ*h, the logarithmic Steenbrink-type vanishing H^q(\\tilde X, Ω^p_{\\tilde X}(log E) ⊗ O(-E) ⊗ \\tilde π*L ⊗ I(\\tilde π*h)) = 0 holds for all p+q > n. A nef-and-big variant is stated as Theorem 1.5. The proof strategy is: construct a positive Q-line bundle L_δ from the given big line bundle using a refined Demailly approximation and the Negativity Lemma (Theorem 3.2); then either use a complete L2 Poincaré-type argument (Theorem 4.1) or, for the compact case, apply the vanishing theorem of Huang–Liu–Wan–Yang [HLWY23] to the positivity of L_δ.","tokens_in":17700,"tokens_out":7216,"duration_ms":70389,"significance":"If the main theorem is correct, it is a substantial generalization: it removes the projectivity/Kähler assumptions from the classical Steenbrink theorem and replaces ampleness by bigness, with multiplier ideal sheaves in the spirit of Nadel's theorem. The paper also contains a useful positivity statement (Theorem 3.2 / Theorem 1.4) for Q-line bundles with controlled fractional multiplicities, and Theorem 4.1 provides a detailed L2 vanishing argument on weakly pseudoconvex spaces. These are genuine contributions. However, the proof of the central compact-case theorem is not fully justified as written, and the paper relies heavily on the author's own unpublished preprints ([Wat24], [Wat26a], [Wat26b], [Wat26c]) for key ingredients, which are not stated or proved in the manuscript.","major_comments":[{"comment":"The displayed argument applies [HLWY23, Theorem 1.1] to the Q-line bundle L_δ = O(-E) ⊗ \\tilde π*L ⊗ I(\\tilde π*h) ⊗ O(∑_{j∈J}(1-δ_j)E_j), with 1-δ_j ∈ (0,1]. The cited theorem, as used in the Introduction and elsewhere, is a vanishing theorem for holomorphic line bundles, not for Q-line bundles with fractional normal-crossing twists. The paper does not prove that [HLWY23, Theorem 1.1] extends to such Q-line bundles, and clearing denominators does not directly yield the desired vanishing for the untwisted sheaf. Thus the short proof of Theorem 1.3 does not establish the stated vanishing. The theorem may be recoverable by invoking the L2 argument of Theorem 4.1, or by proving the needed Q-line-bundle version of [HLWY23, Theorem 1.1], but that step is missing and is load-bearing.","section":"§4, Proof of Theorem 1.3"},{"comment":"The proof of Theorem 4.1 begins by appealing to [Wat26b, Theorem 1.5], [Wat26c, Theorem 4.17], and [Wat25b, Lemma 3.2] for the existence of the quasi-plurisubharmonic function ψ and the equality I(μ*h)=I(H). These results are not stated in the present paper, and several of the cited preprints are unpublished arXiv manuscripts. Since Theorem 4.1 is the main vehicle for the weakly pseudoconvex case and also the fallback justification for Theorem 1.3, the reader cannot verify the proof without consulting external sources that may not be available in final form. Please restate the precise statements used, or include proofs of the relevant implications.","section":"§4, Proof of Theorem 4.1 and Theorem 1.3"},{"comment":"The same fractional-twist issue appears in the 'short analytic proof' of Steenbrink's theorem: the proof applies [HLWY23, Theorem 1.1] to A_δ = μ*A ⊗ O(-∑ δ_j E_j), a Q-line bundle with fractional coefficients. The cited theorem is not shown to hold for such Q-line bundles. This does not affect the truth of the classical Theorem 1.1, but it means the paper's new proof of that theorem is not valid as written and needs the same repair as the proof of Theorem 1.3.","section":"§4, Short proof of Theorem 1.1"}],"minor_comments":[{"comment":"There are numerous typos and grammatical slips, e.g., 'resolusion' for 'resolution', 'theor em' for 'theorem', 'independant' for 'independent', and inconsistent hyphenation of 'Kodaira–Akizuki–Nakano'. A careful proofreading pass is needed.","section":"Throughout"},{"comment":"The metric ℏ is defined by setting it to 0 and +∞ on subsets of Exc(μ). Since h is a singular Hermitian metric on a line bundle, the notation 'ℏ(x)=+∞' and 'ℏ(x)=0' is imprecise; the intended meaning should be spelled out in terms of weight functions or local trivializations.","section":"§4, Proof of Theorem 4.1"},{"comment":"The theorem states that the L2-complex is exact but does not explicitly mention the necessary completeness or Nakano-positivity hypotheses on the Poincaré metric; it refers to [HLWY23] for the proof. Please clarify the exact hypotheses under which the logarithmic L2-Dolbeault resolution is exact.","section":"§2.4, Theorem 2.8"},{"comment":"The text says 'By the compactness of V' when V is only relatively compact; it should say 'By relative compactness of V' or 'by compactness of \\overline{V}'.","section":"§3, Proof of Theorem 3.2"}],"recommendation":"major_revision","confidential_remarks":"The author should be asked to address the Q-line-bundle gap explicitly and to make the paper more self-contained. The main theorem is likely correct, and the L2 argument in Theorem 4.1 gives a plausible route to a complete proof, but the manuscript as submitted does not fully establish Theorem 1.3. The heavy dependence on the author's own unpublished preprints, especially for the proof of Theorem 4.1, should be flagged in the review process; the editor may wish to ask for statements or proofs of the imported results."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The main theorem is exactly the kind of generalization that makes sense: replace ample by big, add multiplier ideal sheaf, allow log resolution, and get vanishing in all bidegrees p+q > n. As far as I can tell, this is genuinely new — prior work by the same author only covers top p or top q, not the full range. The machinery built here, especially the refined Demailly approximation that preserves multiplier ideals (Theorem 3.1) and the positivity of Q-line bundles with divisorial Lelong number corrections (Theorem 3.2), is useful in its own right.\n\nThe bigger issue is the proof of Theorem 1.3. After constructing Lδ as O(-E) ⊗ π̃*L ⊗ I(π̃*h) ⊗ O(Σ(1-δ_j)E_j), with coefficients in (0,1], the proof says 'applying [HLWY23, Theorem 1.1]' gives the desired vanishing for the untwisted integral sheaf. That is not justified as written: the cited theorem is for holomorphic line bundles, and Lδ is a Q-line bundle with a fractional snc twist. Clearing denominators would change the bundle, and the paper does not explain why the vanishing survives. The same gap appears in the short analytic proof of the original Steenbrink theorem (Theorem 1.1) in the final section. It's not fatal — the L2 proof in Theorem 4.1 shows a way to get the result without invoking that theorem on a Q-line bundle — but it needs to be fixed before the compact case is established.\n\nTwo more soft spots, both addressable. The paper leans on several unpublished preprints ([Wat24], [Wat26b], [Wat26c], [Wat25b]) for load-bearing statements, which makes independent verification hard. And there are set-theoretic typos in the descent construction in Theorem 4.1, e.g., V \\ Exc(µ) where Exc(µ) lives in \\hat Y, not V. These should be cleaned up.\n\nWho is this for? Anyone working on vanishing theorems, multiplier ideal sheaves, or classification of compact complex spaces. It's a solid contribution if the gap is closed. I'd send it to a serious referee, not desk reject. If I referee it, I'd ask the author to justify the Q-line bundle step or replace it with the L2 argument.","headline":"A natural extension of Steenbrink vanishing to big line bundles, with a plausible main theorem but a real gap in the compact-case proof that should be fixable.","tokens_in":18225,"tokens_out":4736,"would_cite":false,"duration_ms":44147,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["32L20","14F17","32S20","14F18","32J25","32C15"],"pacs":[],"model":"deepseek-v4-flash","headline":"The Steenbrink vanishing theorem extends from ample to big line bundles on compact complex spaces.","keywords":["Steenbrink vanishing theorem","big line bundles","singular Hermitian metrics","multiplier ideal sheaves","logarithmic differential forms","compact complex spaces","Moishezon spaces","L2-Dolbeault resolution"],"falsifier":"Compute $H^q(\\tilde{X}, \\Omega^p_{\\tilde{X}}(\\log E)\\otimes \\mathcal{O}_{\\tilde{X}}(-E)\\otimes \\tilde{\\pi}^*L\\otimes \\mathcal{I}(\\tilde{\\pi}^*h))$ for a concrete big line bundle whose resolution has a nonzero fractional coefficient $\\delta_j$. A natural test case is the blow-up of $\\mathbb{P}^n$ at a point with $\\pi^*\\mathcal{O}(1)$ and the strict transform of a hyperplane, the setting of Counterexample 4.4; a nonzero group with $p+q>n$ would disprove the main theorem, while vanishing would show the fractional twist is harmless.","tokens_in":17194,"feed_emoji":"📐","tokens_out":9240,"duration_ms":82804,"temperature":0.7,"pith_summary":"This paper extends the Steenbrink vanishing theorem, originally for ample line bundles on complex projective varieties, to big line bundles on compact complex spaces of arbitrary dimension, using multiplier ideal sheaves to control the non-ample part of the bundle. The main theorem says that bigness alone forces the underlying space to be Moishezon and, after resolving singularities, kills all logarithmic cohomology groups in degrees $p+q>n$ when the bundle is twisted by $\\mathcal{O}(-E)$ and the multiplier ideal sheaf. Along the way, the paper proves that bigness can be approximated by a metric with algebraic singularities whose divisorial Lelong numbers produce a positive fractional twist of a $\\mathbb{Q}$-line bundle. It also removes projectivity from the hypothesis, treats nef and big line bundles with trivial multiplier ideal, and gives a new short analytic proof of the classical Steenbrink theorem. If the main theorem is right, vanishing theorems that usually require ampleness become available for the much larger class of big line bundles on non-projective spaces.","feed_headline":"Steenbrink vanishing now holds for big line bundles","feed_subtitle":"The extension works on compact complex spaces, adds multiplier ideal twists, and forces such spaces to be Moishezon.","key_machinery":"The machine is a two-step positivity transfer. First, a refined approximation result turns an arbitrary singular positive metric on a big line bundle into one with algebraic singularities without changing the multiplier ideal sheaf. Second, Theorem 3.2 shows that after a log resolution $\\pi$, the $\\mathbb{Q}$-line bundle $\\pi^*L\\otimes \\mathcal{I}(\\pi^*\\hbar)\\otimes \\mathcal{O}(-\\sum_j \\delta_j E_j)$ is positive for rational $\\delta_j\\in[\\upsilon_j-\\lfloor\\upsilon_j\\rfloor,1)$; this is the exact object that lets the proof slide from the fractional twist $\\delta_j$ to the integral twist $1$ in $\\mathcal{O}(-E)$ and feeds the logarithmic $L^2$-Dolbeault machinery, whose curvature commutator $[","core_discovery":"On a compact complex space $X$ of pure dimension $n$, the paper claims that a big line bundle $L$ forces $X$ to be Moishezon and forces a very strong cohomology vanishing once singularities are resolved. More precisely, $L$ admits a singular Hermitian metric $h$ whose pullback under a resolution has algebraic singularities away from the exceptional locus and has strictly positive curvature current; after a log resolution $\\pi: \\tilde{X}\\to \\hat{X}$, the cohomology $H^q(\\tilde{X}, \\Omega^p_{\\tilde{X}}(\\log E)\\otimes \\mathcal{O}_{\\tilde{X}}(-E)\\otimes \\tilde{\\pi}^*L\\otimes \\mathcal{I}(\\tilde{\\pi}^*h))$ vanishes whenever $p+q>n$. The same statement is proved for singular positive metrics on rel","pith_inferences":["Editorial inference: since the proof only needs the positivity of the $\\mathbb{Q}$-line bundle $L_\\delta$, the same vanishing should hold for a whole family of fractional twists; the integer coefficient $1$ in $\\mathcal{O}(-E)$ is just one endpoint of the allowable interval.","Editorial inference: the nef-and-big case suggests that on Moishezon spaces a Kawamata–Viehweg-type statement holds without multiplier ideals, once a metric with all divisorial Lelong numbers below $1$ is chosen.","Editorial inference: the fractional coefficients $\\delta_j$ behave like discrepancies along the exceptional divisors, so Theorem 1.4 may connect to minimal-model-type adjustments on non-projective spaces."],"forward_implications":["Bigness of $L$ on a compact complex space forces $X$ to be Moishezon and yields the full Steenbrink-type vanishing in every bidegree $p+q>n$ after a log resolution.","For nef and big line bundles the multiplier-ideal twist can be made trivial, so the vanishing holds with no multiplier ideal at all.","The same argument works on relatively compact weakly pseudoconvex spaces, so the vanishing is independent of compactness and of any projective embedding.","Theorem 1.4 produces an explicit positivity certificate: the fractional coefficients $\\delta_j$ lie in $[\\upsilon-\\lfloor\\upsilon\\rfloor,1)$, connecting the vanishing to divisorial Lelong numbers.","The original Steenbrink theorem and its higher-direct-image variant obtain a new, purely analytic proof via the same logarithmic $L^2$ technique."],"supporting_citations":[{"why":"The original Steenbrink vanishing theorem for ample line bundles on projective varieties that this paper extends to the big case.","marker":"[Ste85, Theorem 2]"},{"why":"Supplies the logarithmic $L^2$-Dolbeault vanishing theorem for positive line bundles; the final step of the proof of Theorem 1.3 applies it to the $\\mathbb{Q}$-line bundle $L_\\delta$.","marker":"[HLWY23, Theorem 1.1]"},{"why":"The refined approximation theorem that produces algebraic-singularity metrics while preserving the multiplier ideal sheaf; it appears here as Theorem 3.1.","marker":"[Wat24, Theorem 3.2]"},{"why":"Characterizes bigness on compact complex manifolds by the existence of singular positive Hermitian metrics, used to lift bigness to the resolution.","marker":"[MM07, Theorem 2.3.30]"},{"why":"Demailly's characterization of big line bundles via singular Hermitian metrics with positive curvature, a foundational input throughout.","marker":"[Dem90]"},{"why":"Supplies the framework of singular Hermitian metrics on complex spaces and the bigness/singular-positivity relation used for non-projective $X$.","marker":"[Wat26b]"},{"why":"Previous extension of top-degree and Bogomolov-type logarithmic vanishing for big line bundles, which Theorem 1.3 strengthens to all bidegrees.","marker":"[Wat26a, Corollary 1.3]"},{"why":"The negativity lemma used to adjust positivity along exceptional divisors so that the fractional twist can be chosen positive.","marker":"[Kaw24, Remark 1.6.2 (2)]"}],"fun_headline_variants":["Steenbrink vanishing now works for big line bundles","Big line bundles satisfy Steenbrink vanishing","Vanishing for big line bundles on compact spaces","Generalized Steenbrink theorem for big line bundles"],"cache_read_input_tokens":2816,"weakest_assumption_plain":"The load-bearing premise is that a positive $\\mathbb{Q}$-line bundle with fractional simple-normal-crossing coefficients still forces the same logarithmic cohomology vanishing as an honest positive line bundle; the entire compact-case proof is a transfer from a fractional twist to the integral bundle $\\mathcal{O}(-E)$.","fun_headline_variants_meta":{"raw":{"variants":["Steenbrink vanishing now works for big line bundles","Big line bundles satisfy Steenbrink vanishing","Vanishing for big line bundles on compact spaces","Generalized Steenbrink theorem for big line bundles"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000192,"raw_usage":{"total_tokens":1098,"prompt_tokens":572,"completion_tokens":526,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":316,"completion_tokens_details":{"reasoning_tokens":463}},"tokens_in":316,"tokens_out":526,"duration_ms":5043,"temperature":1.0,"reasoning_tokens":463,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T00:04:50.562277+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute $H^q(\\tilde{X}, \\Omega^p_{\\tilde{X}}(\\log E)\\otimes \\mathcal{O}_{\\tilde{X}}(-E)\\otimes \\tilde{\\pi}^*L\\otimes \\mathcal{I}(\\tilde{\\pi}^*h))$ for a concrete big line bundle whose resolution has a nonzero fractional coefficient $\\delta_j$. A natural test case is the blow-up of $\\mathbb{P}^n$ at a point with $\\pi^*\\mathcal{O}(1)$ and the strict transform of a hyperplane, the setting of Counterexample 4.4; a nonzero group with $p+q>n$ would disprove the main theorem, while vanishing would show the fractional twist is harmless.","supporting_citations":[],"review_version":1}