{"id":"f8191b7d-3714-4ac7-888d-1f7f3b387640","arxiv_id":"2411.17268","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Adapted reflexive 1-forms on klt Campana orbifolds extend to regular 1-forms on log resolutions of adapted covers, and induce pull-back morphisms for orbifold differentials.","lead":"On mildly singular orbifolds, differential forms with fractional poles, defined by pulling back to a ramified cover, are shown to extend to regular forms on any resolution of the cover. This generalizes the known extension theorem for klt spaces and yields pull-back maps for orbifold differentials used in birational classification.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 24's use of [KS21, Prop. 6.4] is the load-bearing step; if its support estimate fails in the ambient ball, the codim-3 reduction in Theorem 1 collapses.","rationale":"I agree with the reader's identification of Lemma 24 as the weakest assumption. The overall architecture is coherent: Lemma 22 is a plausible reduction to a smooth perfectly adapted cover, Lemma 15 provides such a cover after removing codimension 3, and Lemma 21 is a standard descent argument. I did not find an internal algebraic contradiction. The genuine load-bearing unverified input is the analytic extension lemma, whose proof rests on Saito's decomposition and [KS21, Prop. 6.4] without spelling out the hypotheses of that proposition. The point where this matters is the first sentence of the proof of Theorem 1: 'Lemma 24 allows us to remove a closed subset of codimension at least 3 from C.' If Lemma 24 were false or misapplied, there is no fallback argument in the paper. This is not a claim of error; rather, the main theorem inherits exactly the validity and applicability of that cited result. The reader's conditional verdict is appropriate, and I see no reason to move it. The related concern about novelty relative to [Núñ23] is not a correctness issue and does not change the verdict.","tokens_in":35,"tokens_out":27714,"duration_ms":411180,"concrete_test":"Locate [KS21, Prop. 6.4], transcribe its hypotheses verbatim, and check whether the complex E^d constructed in the proof of Lemma 24 satisfies them exactly for (n,d)=(3,1). Concretely, recompute the required bound for A ∩ Supp(Ext^q(E^1,ω_U)) at a 3-fold with an isolated singularity, comparing codim_B(A)=3 with the threshold demanded by Prop. 6.4. If the threshold is 3, the reduction goes through; if the threshold is 4, or the inequality is ≤ −(q+3) instead of ≤ −(q+2), the proof of Theorem 1 must supply an alternative extension argument across the deleted set.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The proof of Theorem 1 depends on deleting a codimension-3 subset A⊂C and then applying Lemma 15 and Lemma 22; the deletion is harmless only because Lemma 24 guarantees that sections of π_*Ω^1_{\\tilde C} extend uniquely across A. Lemma 24 is therefore load-bearing, and its proof is not self-contained: it invokes Saito's Decomposition Theorem to write Rν_*Ω^d ≅ E^d ⊕ F^d and then applies [KS21, Prop. 6.4] to E^d. The paper verifies two numerical properties — H^k(E^d)=0 for k ≥ n−d+1 and RHom_U(E^d,ω_U^•) ≅ E^{n−d}[n] — and derives a support estimate for Ext^q, but it never states the exact hypotheses of Prop. 6.4. In particular, for q ≤ d−n the argument uses only dim(A) ≤ n−d−2; whether this matches the threshold required by Prop. 6.4, with codimension measured in U rather than in B and with the correct strict or non-strict cutoff, is not demonstrated. If Prop. 6.4 needs codim_U(A) ≥ d+3, or if the F^d summand contributes cohomology in degrees other than H^0, the codim-3 reduction in Theorem 1 is unjustified and the proof would need a different mechanism.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper defines adapted reflexive differentials for Campana orbifolds (C-pairs) with respect to an adapted quasi-cover, and proves Theorem 1: if (X,Δ) is a klt C-pair and γ:C→X is an adapted morphism, then for any log resolution π:C̃→C, pull-back of rational differentials induces a morphism Ω^[1]_{(X,Δ,γ)} → π_*Ω^1_{C̃}. The proof combines three ingredients: a Flenner-type extension lemma (Lemma 24) for sections of π_*Ω^d across codimension-(d+2) subspaces, a local construction of perfectly adapted smooth covers off a codimension-3 locus (Lemma 15), and a reduction to logarithmic differentials via a fiber-product argument (Lemma 22), with Lemma 24 used to delete a codimension-3 subset from the cover before applying the local construction. The paper also derives a pull-back morphism for adapted differentials (Theorem 26).","tokens_in":19793,"tokens_out":10999,"duration_ms":106433,"significance":"If the proof is fully justified, Theorem 1 is a meaningful extension of Kebekus–Rousseau's adapted-differential results to singular covers and singular bases, and it complements the classical extension theorems of GKKP11 by allowing the adapted cover to be worse than klt. The argument is well structured, the statements are precise, and the strategy of removing a codimension-3 set and then using a local perfectly adapted smooth cover is natural and potentially useful. The paper also benefits from making explicit the analytic ingredient in Lemma 24 and from stating the pull-back application in Theorem 26. However, the proof of Lemma 24 is a sketch that invokes a deep result, [KS21, Prop. 6.4], without checking its exact hypotheses; since the entire proof of Theorem 1 relies on Lemma 24, this is a load-bearing gap that requires attention.","major_comments":[{"comment":"The proof of Lemma 24 invokes [KS21, Proposition 6.4] but never states the hypotheses of that proposition. The verification only checks two numerical properties of the direct summand E^d (vanishing of H^k(E^d) for k ≥ n−d+1 and the duality RHom(E^d,ω_U^•) ≅ E^{n−d}[n]) and then derives the support estimate dim(A∩Supp(Ext^q(E^d,ω_U^•))) ≤ −(q+2) for all q. It is not shown that this is precisely the condition required by Prop. 6.4, nor is it clarified whether the codimension threshold in Prop. 6.4 is measured in the ambient ball U or in the subspace B, and whether the inequality is strict or non-strict. Since Theorem 1 uses Lemma 24 to justify deleting a codimension-3 subset from the cover C, if the support estimate is not the correct hypothesis, the main proof collapses. The authors should either state Prop. 6.4 verbatim and verify its hypotheses in detail, or replace Lemma 24 by a direct citation of a theorem in [KS21] (or Flenner's theorem) that implies the needed extension statement.","section":"§3, Lemma 24"}],"minor_comments":[{"comment":"The typesetting of the arXiv version contains numerous encoding artifacts (for example, the header 'A/b.sc/t.sc/r.sc/a.sc/c.t.sc/t.sc.' and similar dotted abbreviations in section titles); these should be corrected in a final version.","section":"Throughout"},{"comment":"In the sentence 'And since C is smooth and γ_0^*⌊Δ⌋ is snc, the latter is just the locally free sheaf Ω^d_{C_0}(log⌊Δ⌋)', the symbol C should be C_0; the current wording is confusing because C is not introduced in that part of the proof.","section":"§3, proof of Lemma 22"},{"comment":"The sentence 'Since γ is a quasi-cover, this implies that we may also remove a closed subset of codimension at least 3 from B' would benefit from a short justification: for a quasi-finite dominant morphism between varieties of the same dimension, the closure of the image of a closed subset of codimension at least 3 again has codimension at least 3.","section":"§3, proof of Theorem 1"},{"comment":"The statement of Lemma 24 uses codim_B(A) for a closed complex subspace A of a possibly singular reduced complex space B; the codimension should be defined explicitly (e.g., codim_B(A) = dim B − dim A, with the convention that the intersection is taken over irreducible components).","section":"§3, Lemma 24"},{"comment":"The note says that most contents are extracted from the author's PhD thesis [Núñ23] and refers there for more detailed computations; for the journal version, the authors should consider expanding the proofs of the key lemmas (especially Lemma 24) so that the paper is more self-contained, or at least mark exactly which arguments are deferred to the thesis.","section":"§1.3"}],"recommendation":"major_revision","confidential_remarks":"The paper is a compact research note largely drawn from the author's thesis. The main theorem is interesting and the overall strategy is sound, but the proof of Lemma 24 is not self-contained and the precise use of [KS21, Prop. 6.4] needs to be made explicit. Please ensure the revised version either proves Lemma 24 directly or states the cited result and verifies its hypotheses. The dependence on an unpublished thesis is acceptable in this field but should be minimized."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a genuinely useful condensed write-up of the author's thesis result, and the main theorem appears correct, but it is not new relative to the cited thesis, and the proof leans on a nontrivial analytic input (Lemma 24) that is not fully self-contained. I would send it to a serious referee if the venue accepts thesis-extract notes; otherwise it is a borderline desk reject.\n\nThe paper's content: Theorem 1 is the extension theorem for adapted reflexive 1-forms on klt C-pairs, generalizing Kebekus-Rousseau to singular bases. That is a real result. The proof strategy is sensible: remove a codimension-3 set, use the Flenner/KS21 extension lemma, then apply local perfectly adapted covers and compare adapted with logarithmic differentials. The exposition is clear, and the local computations in Remark 17 are helpful. Lemmas 21 and 22 are clean reductions that I did not find any issue with.\n\nSoft spots: the main one is Lemma 24, as both the reader and the stress-test note. The proof is a sketch: it invokes Saito's decomposition and [KS21, Prop. 6.4] but does not state the exact hypotheses of that proposition. That said, on reading, the numerical conditions verified in the proof are precisely the support estimates needed for extension across codimension d+2 sets: for q ≤ d−n they use dim(A) ≤ n−d−2, and for q ≥ d−n+1 they get vanishing from the decomposition properties. So I do not think the stress-test concern collapses the proof; it is a matter of writing out the cited criterion more explicitly. A referee should ask for that.\n\nSecond: Section 1.3 states plainly that most contents are extracted from [Núñ23], and several lemmas point to the thesis for details. That makes the novelty of this arXiv note low. But it is not a flaw if the venue accepts thesis extracts; it is a disclosure issue, and the author does disclose. The citation pattern is fine: [KR24], [GKKP11], [KS21], [CP19] are all appropriate, and the thesis is self-cited with DOI, which is legitimate.\n\nWho it is for: anyone working on Campana orbifolds, adapted differentials, or extension theorems in the MMP. It would be a good reading-group paper to get the statement and main ideas, but not to verify every step.\n\nRecommendation: if the journal considers thesis-extract notes, send to a competent referee with a request to check Lemma 24 carefully and ask for self-contained statements. If the venue prizes novelty over exposition, desk reject. My own verdict: worth one serious referee round, mostly because the result is useful and correct-looking.","headline":"A clean, honest write-up of a thesis result: the main extension theorem looks right, but the paper is explicitly an extract and Lemma 24 needs to be spelled out before I'd fully trust it.","tokens_in":20440,"tokens_out":2912,"would_cite":true,"duration_ms":27108,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14E30","14F10","14J17","14B05"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that on a klt geometric orbifold, every adapted reflexive 1-form on a ramified cover extends to a regular 1-form on any log resolution of that cover.","keywords":["adapted differentials","geometric orbifolds","C-pairs","klt singularities","reflexive forms","extension theorems","log resolutions","pull-back morphisms"],"falsifier":"Look for a klt C-pair $(X,\\Delta)$ and an adapted cover $\\gamma:C\\to X$ such that some adapted local generator (for instance $y^{E/A-1}dy$ in the coordinates of Remark 17) pulls back to a 1-form on a log resolution $\\widetilde C$ with a pole along an exceptional divisor. A single explicit pole, or a single reduced complex space of pure dimension $n$ with a closed subspace of codimension $d+2$ across which a section of $\\pi_*\\Omega^d$ fails to extend, would contradict Theorem 1 through Lemma 24.","tokens_in":19312,"feed_emoji":"📐","tokens_out":9130,"duration_ms":76918,"temperature":0.7,"pith_summary":"This paper proves an extension theorem for adapted differential forms on klt geometric orbifolds (C-pairs). Such an orbifold is a normal variety $X$ together with a $\\mathbb{Q}$-divisor $\\Delta$ whose coefficients record fractional pole orders $(m_i-1)/m_i$; an adapted morphism is a finite cover on which the orbifold divisor becomes integral. The main theorem says that for a klt pair $(X,\\Delta)$, every adapted reflexive 1-form on the covering space extends to a regular 1-form on any log resolution of that cover. The point is that the cover itself may have worse-than-klt singularities, so ordinary extension of all reflexive forms fails; the adapted ones still extend. This matters because adapted differentials are the orbifold replacement for regular Kähler differentials, and their extension behaviour underlies pull-back maps and birational arguments in orbifold classification.","feed_headline":"Klt orbifolds extend adapted 1-forms to all resolutions","feed_subtitle":"Even when the cover is worse than klt, its adapted differential forms stay regular on the resolution.","key_machinery":"The central object is the sheaf of adapted reflexive differentials $\\Omega^{[1]}_{(X,\\Delta,\\gamma)}$, defined as the reflexive hull of the kernel of a residue morphism on $\\gamma^*\\Omega^1_X(\\log\\lceil\\Delta\\rceil)$: over the snc locus it consists of local forms with pole orders prescribed by the orbifold coefficients, such as $y_i^{E_i/A_i-1}dy_i$ in suitable coordinates. The proof is carried by three mechanisms. Lemma 15 shows that off a codimension-3 subset, a klt C-pair admits analytic-local perfectly adapted covers from smooth varieties; Lemma 22 shows that once such a smooth perfect cover exists, extension on any adapted cover follows by pulling back to a fiber product and invoking logarithmic differential facts; Lemma 24, a decomposition-theory statement, proves that sections of $\\pi_*\\Omega^d_{\\widetilde B}$ extend uniquely across closed subspaces of codimension at least $d+2$, which lets the codimension-3 reduction in Theorem 1 be made. Lemma 20, the invariance of adapted differentials under changing the adapted cover, ties the pieces together.","core_discovery":"On the paper's own terms, the discovery is Theorem 1: if $(X,\\Delta)$ is a klt C-pair and $\\gamma:C\\to X$ is an adapted morphism, then for any log resolution $\\pi:\\widetilde C\\to C$ the pull-back of rational differential forms induces a morphism of $\\mathcal O_C$-modules $\\Omega^{[1]}_{(X,\\Delta,\\gamma)}\\to \\pi_*\\Omega^1_{\\widetilde C}$. In words, every adapted reflexive 1-form on $C$ is regular after pulling back to a resolution of $C$, even though $C$ itself may be far from klt. The proof reduces the statement to a local model: outside a codimension-3 set, the pair admits a perfectly adapted cover from a smooth variety, and on that model adapted forms coincide with logarithmic forms on the smooth cover, which are known to extend after pulling back and resolving. A second ingredient removes the codimension-3 defect by showing that sections of $\\pi_*\\Omega^d_{\\widetilde C}$ extend uniquely across closed subspaces of codimension at least $d+2$. The paper also derives a pull-back theorem for adapted 1-forms analogous to the classical klt pull-back.","pith_inferences":["Editorial extension: one could test whether the same statement holds for all degrees $d$ on arbitrary klt C-pairs without the smooth-perfect-cover assumption; the paper proves degree 1, and the natural obstruction is whether the codimension-$(d+2)$ extension lemma can be combined with a codimension-3 adapted-cover model for higher-degree adapted forms.","A further testable consequence is that adapted symmetric powers or tensor powers of $\\Omega^{[1]}_{(X,\\Delta,\\gamma)}$ should also extend on log resolutions; if true, this would give orbifold analogues of plurigenera-type or vanishing statements for covers.","One could also ask whether the statement extends to lc C-pairs when the adapted cover is required to be klt; the paper's mechanism suggests the klt assumption on the pair is used exactly to produce perfectly adapted covers off codimension 3."],"forward_implications":["The extension theorem gives a pull-back morphism $\\Omega^{[1]}_{(X,\\Delta,\\gamma)}\\to \\varphi_*\\Omega^{[1]}_D$ for every morphism $\\varphi:D\\to C$ from a normal variety whose image is not contained in the singular locus of $C$ (Theorem 26).","Combined with Lemma 22, the argument proves extension to logarithmic differentials for every degree $d$ whenever the base admits a smooth perfectly adapted cover, in particular for all adapted covers over C-pairs with quotient singularities (Corollary 23).","Because the conclusion is independent of the chosen adapted cover (Lemma 20), the regular extension property is intrinsic to the klt C-pair, not to a particular cover.","The result confirms that the singularities of the cover do not obstruct adapted forms: even if $C$ is worse than klt, the adapted submodule of its reflexive differentials still has no poles on any resolution."],"supporting_citations":[{"why":"Supplies the klt extension theorem and pull-back machinery that Theorem 1 adapts to orbifold covers; also used in Lemma 15 to pass to Q-factorial and quotient-singularity models.","marker":"[GKKP11]"},{"why":"Is the smooth-case extension result for adapted differentials that Theorem 1 generalizes to singular klt pairs.","marker":"[KR24]"},{"why":"Provides Proposition 6.4, the support estimate used in Lemma 24 to extend sections of $\\pi_*\\Omega^d$ across codimension $d+2$ subspaces.","marker":"[KS21]"},{"why":"Supplies the logarithmic differential pull-back and pole order facts used in Lemma 22 to keep forms logarithmic after resolving.","marker":"[GKK10]"},{"why":"Is the classical extension theorem for differential forms on nonisolated singularities whose strategy motivates the codimension argument in Lemma 24.","marker":"[Fle88]"},{"why":"Provides the definition of klt singularities, the cyclic cover construction used in Lemma 12, and the ramification formula used in Lemma 14.","marker":"[KM98]"},{"why":"Gives the definition and local generators of adapted differential forms, the object whose extension is studied.","marker":"[CP19]"},{"why":"Defines C-pairs and their singularities, the setting in which the main theorem is stated.","marker":"[Cam11]"}],"fun_headline_variants":["Adapted forms on klt orbifolds extend to resolutions","Even non-klt covers yield regular forms on resolutions","Adapted differentials regular after pullback on klt orbifolds","Klt orbifolds: adapted 1-forms regular on all resolutions","Klt orbifold adapted forms regular on cover resolutions"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument rests on the assertion that a section of $\\pi_*\\Omega^d$ defined off a closed subspace of codimension at least $d+2$ automatically extends over that subspace; if that extension statement were false, the proof's reduction to the smooth adapted-cover model would collapse.","fun_headline_variants_meta":{"raw":{"variants":["Adapted forms on klt orbifolds extend to resolutions","Even non-klt covers yield regular forms on resolutions","Adapted differentials regular after pullback on klt orbifolds","Klt orbifolds: adapted 1-forms regular on all resolutions","Klt orbifold adapted forms regular on cover resolutions"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001444,"raw_usage":{"total_tokens":5768,"prompt_tokens":846,"completion_tokens":4922,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":462,"completion_tokens_details":{"reasoning_tokens":4834}},"tokens_in":462,"tokens_out":4922,"duration_ms":30479,"temperature":1.0,"reasoning_tokens":4834,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T12:21:28.089520+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Look for a klt C-pair $(X,\\Delta)$ and an adapted cover $\\gamma:C\\to X$ such that some adapted local generator (for instance $y^{E/A-1}dy$ in the coordinates of Remark 17) pulls back to a 1-form on a log resolution $\\widetilde C$ with a pole along an exceptional divisor. A single explicit pole, or a single reduced complex space of pure dimension $n$ with a closed subspace of codimension $d+2$ across which a section of $\\pi_*\\Omega^d$ fails to extend, would contradict Theorem 1 through Lemma 24.","supporting_citations":[],"review_version":1}