{"id":"4943e888-7169-4df2-a5db-849d8f98fffe","arxiv_id":"2506.00873","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper proves that orbifold entire curves on projective varieties with numerically parallel boundary divisors and sufficiently high multiplicity must be algebraically degenerate, a new case of Campana's orbifold conjecture.","lead":"This paper proves a version of Campana's orbifold conjecture for projective varieties with numerically parallel boundary divisors, showing that entire holomorphic curves with high ramification must be algebraically degenerate. A generalist should care because it moves a central conjecture in the study of entire curves on algebraic varieties from special toric cases to general varieties, linking complex analysis and number theory.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 9's inference that numerical equivalence implies linear equivalence after q(X)=0 is unjustified: NS(X) can have torsion, so the morphism φ to G_m^n is not established.","rationale":"The central claim of the paper is a substantial case of Campana's orbifold conjecture, and the overall strategy—reducing to the toric case via a morphism to G_m^n and applying the Ru–Wang toric theorem—is promising. However, the proof's key bridge in Lemma 9 is logically broken: q(X)=0 gives Pic(X)≅NS(X) but not torsion-freeness of NS(X), so numerically parallel divisors need not be linearly parallel. An Enriques surface demonstrates that the required promotion from numerical to linear equivalence can fail exactly in the q=0 setting: the 2-torsion canonical class is numerically trivial but not principal, and one can choose effective divisors in classes L and L+η that are numerically parallel but not linearly equivalent. Without the linear equivalences, the rational functions φ_i used to define the morphism to G_m^n do not exist, and the entire reduction to Theorems 8 and 10 collapses. This is the same load-bearing concern the reader identified in the weakest_assumption field, so I agree with the reader's analysis. The concern is internal to the proof rather than a disagreement with external consensus, and it is concrete: a specific torsion counterexample of the claimed inference. I do not see a reason to move the verdict beyond CONDITIONAL: the gap is real but plausibly repairable by adding a torsion-free assumption on NS(X) or by replacing X with a finite étale cover that trivializes the torsion, and the theorem itself may still be true. Thus the reader's conditional verdict remains appropriate.","tokens_in":9816,"tokens_out":17499,"duration_ms":180368,"concrete_test":"Let X be an Enriques surface (q(X)=0, NS(X) has a 2-torsion class η, e.g. the canonical class). Choose an ample line bundle L with h^0(X,L)≥3; pick general effective divisors D_1,D_3∈|L| and D_2∈|L+η| so that D_1+D_2+D_3 has simple normal crossings and no support of any D_i is contained in the union of the other two. Then d_jD_i≡d_iD_j holds with d_i=1 because η is numerically trivial, but D_2−D_1 is not linearly equivalent to 0. Verify that this triple satisfies all the standing hypotheses of Lemma 9 up to the disputed inference (except the assumed existence of a nondegenerate orbifold curve). If so, the lemma's step 'numerical equivalence implies linear equivalence' is invalid as stated, and the proof of Lemma 9 must be modified to exclude torsion or to kill it via a finite étale cover before the morphism φ is constructed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Located in Section 3, proof of Lemma 9, immediately after 'we have Pic(X) ≅ NS(X)'. The paper argues that because q(X)=0 gives Pic(X)≅NS(X), the numerical equivalences d_jD_i ≡ d_iD_j imply linear equivalences d_jD_i ∼ d_iD_j. This is false in general. Pic^0(X)=0 only identifies Pic(X) with NS(X); it does not make NS(X) torsion-free. Numerical equivalence corresponds to equality in the torsion-free quotient NS(X)/tors, whereas linear equivalence requires equality in NS(X) itself. A non-zero torsion class in NS(X) is numerically trivial without being linearly trivial. Concrete examples exist: Enriques surfaces have q(X)=0 and a nontrivial 2-torsion canonical class that is numerically trivial. Under the paper's assumptions, one can draw effective divisors D_1,D_3 from a large linear system |L| and D_2 from |L+η| with η a torsion class; these are numerically parallel with coefficients d_i=1, but D_2−D_1 is not principal. Therefore the constructed rational functions φ_i with div(φ_i)=d_{n+1}D_i−d_iD_{n+1} need not exist, and the dominant morphism φ:X\\D→G_m^n is unsupported. Since the reduction to the toric theorems (Theorem 8 and Theorem 10) depends entirely on this morphism, the proof of Theorem 1 has a genuine gap at this step. The reader's weakest_assumption identifies exactly this point, and I agree. The gap may be repairable by adding a torsion-freeness hypothesis on NS(X) or by passing to a finite étale cover that kills torsion, but as written the lemma is not proved.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves a special case of Campana's orbifold conjecture: on a smooth projective variety X of dimension n, if D_1,...,D_{n+1} are Z-linearly independent effective divisors whose sum D has simple normal crossings, which are numerically parallel and such that (X,D) is of log general type, then every orbifold entire curve with sufficiently large prescribed multiplicities along the D_i is algebraically degenerate. The proof builds an orbifold version of the logarithmic Bloch-Ochiai theorem, reduces the non-degeneracy assumption to the existence of a dominant morphism from X\\D to G_m^n, and then applies a toric second-main-theorem type result of Ru and Wang. The central claim is a genuine generalization of the earlier projective-space theorem, but the proof as written contains two load-bearing gaps.","tokens_in":10149,"tokens_out":30573,"duration_ms":325374,"significance":"If the result can be made rigorous, it is a meaningful step toward Campana's orbifold conjecture for numerically parallel divisors on general projective varieties, a case not covered by the existing toric methods. The paper is transparent about its reliance on prior results, especially Noguchi-Winkelmann's Nevanlinna theory and the authors' own toric theorem, and the structural reduction from a general variety to G_m^n is a natural and potentially reusable strategy. The main novelty is the combination of numerical parallelism with the quasi-Albanese and Bloch-Ochiai machinery. However, the two gaps identified below currently prevent the central theorem from being accepted as proved.","major_comments":[{"comment":"After deriving q(X)=0, the proof states that Pic(X) ≅ NS(X) and therefore the numerical equivalences d_jD_i ≡ d_iD_j imply the linear equivalences d_jD_i ∼ d_iD_j. This inference is invalid because numerical equivalence is equality in the torsion-free quotient NS(X)/tors, not in NS(X) itself. Since NS(X) can have torsion even when Pic^0(X)=0, the classes d_jD_i - d_iD_j may be non-zero torsion line bundles, which are numerically trivial but not linearly trivial. A concrete obstruction is given by an Enriques surface, which has q=0 and non-trivial 2-torsion; effective divisors D_1,D_3 in one linear system and D_2 in a system twisted by a torsion class are numerically equivalent but not linearly equivalent. The rest of Lemma 9, including the existence of the rational functions φ_i and the morphism to G_m^n, depends on this step. The gap is repairable: if M kills NS(X)_tors, then M(d_jD_i - d_iD_j) is linearly trivial, so one can replace D_i by M D_i and run the argument with the scaled divisors, adjusting the multiplicity parameter ℓ accordingly. This repair should be stated explicitly, because as written Lemma 9 is false.","section":"Section 3, proof of Lemma 9"},{"comment":"After Lemma 9, the proof asserts: 'Lemma 9 implies that there is a positive integer ℓ1 such that ... there exists a finite morphism π : X → P^n(C).' Lemma 9 only establishes a dominant morphism φ : X\\D → G_m^n, together with its extension to a rational map arφ : X → P^n whose pullback of the coordinate hyperplanes gives the D_i. A dominant morphism between projective varieties of the same dimension need not be finite: it may contract curves not contained in D. The hypotheses of Theorem 1 do not rule this out. For example, let X be the blow-up of P^2 at a point and let D_1,D_2,D_3 be the preimages of three smooth plane curves of degree d≥2 that avoid the blown-up point and meet transversally. Then D is normal crossing, the D_i are numerically (indeed linearly) equivalent, and (X,D) is of log general type, yet the morphism defined by |D_i| factors through the blow-down and contracts the exceptional divisor; moreover no finite morphism π : X → P^n can have π^*D0 = D, because D is not ample. Thus the reduction to Theorem 10, which requires a finite morphism, is not justified. The proof should either prove finiteness of the constructed morphism from the hypotheses, or replace this step by a different reduction, for example passing to the Stein factorization and applying the toric theorem to the finite part, or applying the projective-space version of Ru-Wang's theorem directly to the image of f under the dominant morphism.","section":"Section 3, proof of Theorem 1"}],"minor_comments":[{"comment":"The terminology 'Z-linearly independent divisors' is nonstandard: the definition given in the introduction is support-theoretic, not the usual linear independence in Div(X). Although the two notions agree for effective divisors under the stated support condition, the equivalence should be stated and proved or the terminology changed.","section":"Throughout"},{"comment":"The sentence 'φ∗(Hi) = Di+1, for 0 ≤ i ≤ n + 1' should read 'for 0 ≤ i ≤ n', since there are n+1 coordinate hyperplanes and n+1 divisors.","section":"Section 3, proof of Theorem 1"},{"comment":"The phrase 'these divisors are in general position' is used in the statement of Lemma 9 but never defined; the main theorem assumes D is normal crossing, which is not the same as general position for arbitrary effective divisors. This should be clarified.","section":"Lemma 9"},{"comment":"When the proof replaces D_i by (d/d_i)D_i with d = lcm{d_1,...,d_{n+1}}, the new divisors are no longer the original D_i, and this changes the orbifold multiplicities and the integer ℓ. The reuse of the same symbol D_i is confusing; the scaling should be tracked explicitly through the rest of the argument.","section":"Section 3, proof of Lemma 9"},{"comment":"The phrase 'ramification divisor of π omitting components from the support of H' is ambiguous: it should say explicitly that R is the part of the ramification divisor whose support is disjoint from H, and the later sentence in the proof of Theorem 1 about the support of D should be reworded so that H = π^*D0, rather than R, is the divisor containing D.","section":"Theorem 10"}],"recommendation":"major_revision","confidential_remarks":"The paper depends on the authors' own Theorem 10 for the final degeneration step. This is a legitimate dependency rather than circular reasoning, but the editors may wish to ensure that Theorem 10 has been independently refereed and is not under simultaneous submission with unresolved issues. The numerical-to-linear equivalence gap in Lemma 9 is the most serious technical issue; the proposed repair by killing torsion is natural but changes the proof substantially. The finiteness gap in Theorem 1 is also significant and may require a different reduction strategy. I would be willing to review a revised version."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things about this paper: it proves a genuinely new case of Campana's orbifold conjecture for arbitrary projective varieties, and it has a real gap in Lemma 9 that is likely repairable but needs to be fixed before the theorem is established.\n\nWhat is new: the main theorem extends the authors' earlier P^n and toric results to numerically parallel divisors on any smooth projective variety, under log general type. Lemma 9 is the key new mechanism: it promotes numerical parallelism to linear parallelism and builds a dominant morphism to G_m^n, reducing to the toric case. The orbifold version of the Bloch-Ochiai theorem (Theorem 5) is also a useful contribution.\n\nThe paper does a lot well. The structure is clear, the reduction strategy is sensible, and the reliance on prior machinery (Noguchi-Winkelmann, Huang-Levin, and their own [11]) is explicit. The self-citation is a genuine dependency, not circularity: the final degeneration step uses their earlier toric theorem, which is the right tool given the reduction.\n\nThe soft spot: in Lemma 9, after deriving q(X)=0, the authors conclude that the numerical equivalences d_j D_i ≡ d_i D_j imply linear equivalences from Pic ≅ NS. That is not valid. Numerical equivalence only gives equality in NS modulo torsion. If NS has torsion, the difference can be a non-zero torsion class, which is numerically trivial but not linearly trivial. Enriques surfaces already give examples. Without the linear equivalences, the rational functions φ_i need not exist, and the morphism to G_m^n collapses. The gap is load-bearing. It can likely be repaired either by adding a torsion-freeness hypothesis on NS(X) or by passing to a finite étale cover killing the torsion and pulling everything back; but as written the lemma is not proved.\n\nThere is also a minor wording issue in the proof of Theorem 1: the ramification divisor R is said to \"omit components from the support of π^*(D0)\" and then \"contains the support of D,\" which are contradictory since D is exactly that support. This is probably a typo but should be cleaned up.\n\nBottom line: the theorem is significant and the strategy is sound, but the proof as written is incomplete at a key step. It deserves a serious referee—this is exactly the kind of paper that should go to review, with instructions to the authors to fix Lemma 9. I wouldn't cite it in its current form, but I expect a repaired version to be citable.","headline":"A genuinely new case of Campana's orbifold conjecture for general varieties, but the proof of Lemma 9 has a load-bearing gap in passing from numerical to linear equivalence.","tokens_in":10701,"tokens_out":3257,"would_cite":false,"duration_ms":31670,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["32H30","32Q45","30D35"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that on a smooth projective variety with n+1 numerically parallel effective divisors of log general type, every orbifold entire curve with sufficiently high multiplicity is algebraically degenerate.","keywords":["Campana orbifold conjecture","entire curves","algebraic degeneracy","numerically parallel divisors","orbifold divisors","log general type","Nevanlinna theory","semi-abelian varieties"],"falsifier":"Take a smooth projective variety $X$ with $q(X)=0$ and a nonzero torsion class in $NS(X)$ that is numerically trivial but not linearly trivial, and arrange effective divisors $D_1, \\ldots, D_{n+1}$ in normal-crossing position with $d_jD_i \\equiv d_iD_j$ numerically but no common scaling linearly equivalent; if such a log-general-type configuration exists, Lemma 9 cannot produce the dominant map to $\\mathbb{G}_m^n$, so the reduction to the toric theorem collapses.","tokens_in":9585,"feed_emoji":"📐","tokens_out":11225,"duration_ms":98299,"temperature":0.7,"pith_summary":"This paper proves a case of Campana's orbifold conjecture for general projective varieties. The setting is a nonsingular projective variety $X$ of dimension $n$ with $n+1$ effective divisors $D_1, \\ldots, D_{n+1}$ that are $\\mathbb{Z}$-linearly independent, numerically parallel, have normal-crossing sum, and make $(X,D)$ of log general type. The theorem says that for any orbifold divisor with these components as support, once the required multiplicity along each component is taken large enough, every entire curve into the orbifold is algebraically degenerate. This extends earlier results from projective space to arbitrary varieties, using numerical parallelism to construct a dominant map to a torus and then applying the toric case.","feed_headline":"High-multiplicity orbifold curves must be algebraically degenerate","feed_subtitle":"A new case of Campana's orbifold conjecture: past a large enough multiplicity, entire curves must be degenerate.","key_machinery":"The central machinery consists of orbifold divisors $\\Delta = \\sum_i (1-m_i^{-1})D_i$, the multiplicity threshold $\\ell$, and two transfer theorems: the orbifold logarithmic Bloch-Ochiai theorem, which converts high multiplicity into first-main-theorem growth estimates, and a toric degeneration theorem for entire curves highly ramified over the boundary. Inside the proof, Lemma 9 is the load-bearing step: it uses a hypothetical nondegenerate orbifold curve to force $q(X)=0$, converts numerical equivalence into linear equivalence, and constructs the dominant map $\\phi: X \\setminus D \\to \\mathbb{G}_m^n$. That map transfers the whole problem to a toric variety, where the toric case supplies the contradiction.","core_discovery":"The paper's central claim is that numerical parallelism plus the log-general-type condition forces rigidity: if $f: \\mathbb{C} \\to (X, \\Delta)$ is an orbifold entire curve with multiplicity at least some integer $\\ell$ along each $D_i$, then $f(\\mathbb{C})$ lies in a proper algebraic subvariety of $X$. The proof shows that a hypothetical nondegenerate curve would force $q(X)=0$ and would promote numerical equivalences $d_jD_i \\equiv d_iD_j$ to linear equivalences $d_jD_i \\sim d_iD_j$, yielding a dominant morphism $X \\setminus D \\to \\mathbb{G}_m^n$ that pulls back coordinate hyperplanes to the divisors $D_i$. The orbifold curve then pushes forward to an orbifold curve in projective space, where the toric case of the conjecture gives the contradiction.","pith_inferences":["A testable next step is to check whether the conclusion $q(X)=0$ can be replaced by a rank condition on the subgroup generated by the divisors, which would relax the $\\mathbb{Z}$-linear independence assumption.","The same argument might handle more than $n+1$ numerically parallel divisors by choosing $n+1$ independent ones and treating the remaining components as extra ramification data.","If the numerical-to-linear promotion fails on some variety with torsion in $NS(X)$, the theorem might still be true but would need a new way to produce the torus map; searching for such an example would isolate the real boundary of the method.","Tracking the constants in the Bloch-Ochiai and toric theorems could make the threshold $\\ell$ effective, turning the existence statement into a computable bound."],"forward_implications":["If Theorem 1 is correct, any nondegenerate entire curve into such an orbifold is impossible once the multiplicity threshold $\\ell$ is crossed, so the orbifold is hyperbolic in the strong orbifold sense.","The theorem extends the known toric case to arbitrary nonsingular projective varieties under numerical parallelism, making the earlier projective-space result a special case.","The proof implies that a nondegenerate configuration can only occur when $q(X)=0$ and the normalized divisors are linearly equivalent, which is a structural rigidity statement independent of the degeneracy conclusion.","The result supplies a function-theoretic counterpart to arithmetic statements about integral points on complements of numerically parallel divisors."],"supporting_citations":[{"why":"Supplies the orbifold logarithmic Bloch-Ochiai theorem, jet differential estimates, quasi-Albanese machinery, and the semi-toric subgroup criterion used in Theorems 5 through 8.","marker":"[8]"},{"why":"Provides the toric-variety theorem (Theorem 10) that delivers the final contradiction after the construction of the map to projective space.","marker":"[11]"},{"why":"Its Lemma 3.1 on numerically parallel divisors is the template that Lemma 9 adapts to the orbifold setting.","marker":"[4]"},{"why":"Gives the positivity fact that log general type of $(X,D)$ implies $(X,\\Delta_\\ell)$ is of general type for large enough $\\ell$.","marker":"[5]"},{"why":"Defines orbifold divisors and orbifold entire curves, fixing the objects of the conjecture.","marker":"[1]"},{"why":"Supplies the general-position argument used to verify that the ramification divisor and the boundary are in general position in the proof of Theorem 1.","marker":"[2]"},{"why":"The arithmetic theorem whose complex analogue motivates and underlies Theorem 8's description of Zariski closures in $\\mathbb{G}_m^n$.","marker":"[6]"}],"fun_headline_variants":["High multiplicity forces algebraic degeneration in orbifolds","Campana's conjecture proven for numerically parallel divisors","Numerically parallel divisors imply degenerate entire curves","Multiplicity threshold yields degenerate orbifold curves","Orbifold curves of high multiplicity become algebraically degenerate"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that on a smooth projective variety with $q(X)=0$, numerical equivalence $d_jD_i \\equiv d_iD_j$ can be promoted to linear equivalence $d_jD_i \\sim d_iD_j$; this needs $NS(X)$ to have no non-zero numerically trivial or torsion classes, which the paper does not establish.","fun_headline_variants_meta":{"raw":{"variants":["High multiplicity forces algebraic degeneration in orbifolds","Campana's conjecture proven for numerically parallel divisors","Numerically parallel divisors imply degenerate entire curves","Multiplicity threshold yields degenerate orbifold curves","Orbifold curves of high multiplicity become algebraically degenerate"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000215,"raw_usage":{"total_tokens":1427,"prompt_tokens":939,"completion_tokens":488,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":555,"completion_tokens_details":{"reasoning_tokens":415}},"tokens_in":555,"tokens_out":488,"duration_ms":4570,"temperature":1.0,"reasoning_tokens":415,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T11:56:31.808059+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a smooth projective variety $X$ with $q(X)=0$ and a nonzero torsion class in $NS(X)$ that is numerically trivial but not linearly trivial, and arrange effective divisors $D_1, \\ldots, D_{n+1}$ in normal-crossing position with $d_jD_i \\equiv d_iD_j$ numerically but no common scaling linearly equivalent; if such a log-general-type configuration exists, Lemma 9 cannot produce the dominant map to $\\mathbb{G}_m^n$, so the reduction to the toric theorem collapses.","supporting_citations":[{"cited_title":"Noguchi and J","cited_arxiv_id":null,"evidence_quote":"Supplies the orbifold logarithmic Bloch-Ochiai theorem, jet differential estimates, quasi-Albanese machinery, and the semi-toric subgroup criterion used in Theorems 5 through 8."},{"cited_title":"Ru and J","cited_arxiv_id":null,"evidence_quote":"Provides the toric-variety theorem (Theorem 10) that delivers the final contradiction after the construction of the map to projective space."},{"cited_title":"Lazarsfeld, Positivity in Algebraic Geometry","cited_arxiv_id":null,"evidence_quote":"Gives the positivity fact that log general type of $(X,D)$ implies $(X,\\Delta_\\ell)$ is of general type for large enough $\\ell$."},{"cited_title":"Campana, Orbifolds, special varieties and classification theory , Ann","cited_arxiv_id":null,"evidence_quote":"Defines orbifold divisors and orbifold entire curves, fixing the objects of the conjecture."},{"cited_title":"Guo, C.-L","cited_arxiv_id":null,"evidence_quote":"Supplies the general-position argument used to verify that the ramification divisor and the boundary are in general position in the proof of Theorem 1."},{"cited_title":"Laurent, ´Equations diophantiennes exponentielles , Invent","cited_arxiv_id":null,"evidence_quote":"The arithmetic theorem whose complex analogue motivates and underlies Theorem 8's description of Zariski closures in $\\mathbb{G}_m^n$."}],"review_version":1}