{"id":"7e8c35d9-6d58-49ff-845f-25be33accb00","arxiv_id":"2506.17537","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Birationally admissible families of stable minimal models with maximal variation have log general type bases, yielding stratified hyperbolicity of the KSBA moduli stack.","lead":"This paper proves that the moduli stack of stable minimal models can be divided into finitely many pieces, each of which is hyperbolic: every covering family over a piece forces the base to be of log general type. The result extends previously known hyperbolicity theorems for smooth families to singular boundary families of stable minimal models.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Main theorem is conditional on unverified black boxes: Birkar's Theorem 5.1 (proper DM stack) and Proposition 5.3 (ampleness of λ_{a,r}); if either fails, det(W) is not big and the Viehweg-Zuo construction in Theorem 5.10 has no entry point.","rationale":"The reader's weakest_assumption identifies exactly the same load-bearing point: Theorem 1.1 is founded on Theorem 5.1 and Proposition 5.3, both of which are imported from Birkar's preprint and from Fujino/Kovács-Patakfalvi without independent verification. I agree with this assessment. The internal VZ construction—Borel-Moore variations of mixed Hodge structure, lower canonical extensions, the Higgs subsheaf generated by a big line bundle—is elaborate and appears to follow the Campana-Paun/Popa-Schnell architecture, and the paper gives substantial new technical content in Sections 2–4. The Q-coefficient issue in Definition 2.1/Lemma 2.8 and the typo in (5.16) are real but appear reparable by replacing D with its integral ceiling/support and by correcting pullbacks; they do not by themselves invalidate the central claim. The definitional looseness around strict birationally admissibility over non-smooth strata (Definitions 6.4–6.5) also seems patchable by intersecting with smooth loci. By contrast, a gap in Birkar's moduli theorem or in the ampleness of λ_{a,r} would remove the very ground on which det(W) is big, so it is the single most load-bearing concern. This does not change the reader's conditional verdict: the paper should be accepted only subject to verification of these external foundations.","tokens_in":47016,"tokens_out":22291,"duration_ms":245217,"concrete_test":"Independently trace the two cited inputs: (1) Confirm that Birkar [4, Theorem 1.14] indeed constructs a proper Deligne-Mumford stack M_slc(d, Φ_c, Γ, σ) for the stated class of stable minimal models over Spec(C), with the listed enumerative invariants; (2) verify that [16] and [35] prove nefness of f_*(r(K_{X/S}+B+aA)) for families over smooth projective curves in this generality (slc fibers, K+B semi-ample, A ample, coefficients as in Φ_c), and that Kollár's criterion [27, §2.9] together with Birkar's boundedness data [4, Lemma 10.2] implies λ_{a,r} is ample on the coarse moduli space. If either step fails, recompute Theorem 5.10 Step 3: without ampleness of λ_{a,r}, det(W) is not known big and the inclusion (5.16) cannot be made.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The proof of Theorem 1.1 passes through Theorem 5.10, whose final step needs det(W) to be big on the compactified base eS. This bigness is obtained exactly as (ξ∘σ)^*λ_{a,r} = det(W), with λ_{a,r} ample by Proposition 5.3 and ξ∘σ generically finite. Proposition 5.3 is asserted in two sentences, citing Kollár's criterion [27, §2.9] plus nefness results of Fujino [16] and Kovács-Patakfalvi [35]. The statement needed here is not literally the KSBA projectivity proved in those papers: it concerns Birkar's full class of (d, Φ_c, Γ, σ)-stable minimal models, where K_X+B is only semi-ample and the polarization A is auxiliary, and it requires the determinant line bundle to descend to and be ample on the coarse moduli space of the DM stack. The stack itself comes from Birkar's preprint [4, Theorem 1.14], which is used as a black box and is not independently verified in the manuscript. If either Birkar's completeness/properness theorem or the ampleness of λ_{a,r} has a hidden gap, then det(W) need not be big, the inclusion (5.16) has no basis, and the entire VZ-Higgs construction in Theorem 5.10 collapses, taking Theorems 1.1 and 1.3 with it. This is a dependency risk rather than an internal contradiction, but it is genuinely load-bearing: the paper contributes the elaborate Borel-Moore VMHS and Higgs-sheaf machinery, yet its central conclusion is only as secure as these external inputs.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper claims a stratified hyperbolicity theorem for Birkar's moduli stack of stable minimal models. The main result, Theorem 1.1, states that the base of any birationally admissible family of (d, Φ_c, Γ, σ)-stable minimal models with generically finite classifying map is of log general type; Theorem 1.3 then concludes that each stratum of a birationally admissible stratification is hyperbolic. The proof develops a variation-of-mixed-Hodge-structures package for Borel-Moore cohomology of simple normal crossing pairs, proves admissibility and a geometric description of the lower canonical extension, constructs a Viehweg-Zuo type Higgs sheaf containing the Kollár-type determinant polarization, and uses the Campana-Paun big-cotangent criterion. The main theorems are explicitly conditional on Birkar's properness theorem for the moduli stack and on the ampleness of the determinant line bundle λ_{a,r}, which are used as black boxes.","tokens_in":47128,"tokens_out":15307,"duration_ms":153957,"significance":"If the main theorems are correct, the paper gives a positive answer to a question of Birkar on the global geometry of moduli of stable minimal models, and it extends the Viehweg-Zuo/Campana-Paun/Popa-Schnell hyperbolicity results beyond the smooth principal stratum to the boundary strata of the stack. The stratified statement appears to be new even for the canonical stratification of M_{g,n}. The paper's constructive machinery—Borel-Moore VMHS, admissibility, lower canonical extension, and the fiber-product argument—is substantial and, modulo the external inputs, mostly carried out in detail. The inclusion of the determinant polarization into the top Hodge bundle is an explicit and useful observation. However, the central claim rests on external theorems whose exact hypotheses are not verified, and there is an internal inconsistency between the Q-divisor convention in Definition 2.1 and the line-bundle statements that follow.","major_comments":[{"comment":"Definition 2.1 allows D ≥ 0 to be a Weil Q-divisor with arbitrary nonnegative coefficients, but Lemma 2.8 asserts an isomorphism F^n_{(X,D),BM} ≃ O_X(K_X + D)[−n] and Eq. (3.3) uses O_X(K_{X/S} + D − f^*D_S)[−n]. For a Q-divisor D, the sheaves O_X(K_X + D) and O_X(−D) are not defined as line bundles, and the decompositions D[k], the residue/Gysin maps, and Lemma 2.5 require D to be reduced, or at least integral. Since Section 4 and Theorem 5.10 apply the construction to Q-divisors such as Δ_o = B_o + aA_o and its pullbacks, this is not a harmless convention: the isomorphism in Lemma 2.8 is exactly what produces the inclusion of the determinant line bundle into the top Hodge bundle in Lemma 3.19 and Lemma 4.3. The authors should either restrict Definition 2.1 to integral/reduced divisors and explain how Q-boundaries are converted, or extend the proof of Lemma 2.8 to Q-divisors explicitly.","section":"Definition 2.1; Lemma 2.8; Eq. (3.3)"},{"comment":"The entry point of the Viehweg-Zuo machinery in Theorem 5.10 is the bigness of det(W), obtained as (ξ∘σ)^*λ_{a,r} = det(W) with λ_{a,r} ample by Proposition 5.3 and ξ∘σ generically finite. Proposition 5.3 is proved in two sentences by citing Kollár's criterion [27, §2.9] and nefness results of Fujino [16] and Kovács-Patakfalvi [35]. The statement needed here, however, is not literally the KSBA projectivity proved in those papers: it concerns Birkar's full class of (d, Φ_c, Γ, σ)-stable minimal models, where K_X + B is only semi-ample and A is an auxiliary polarization, and it requires the determinant line bundle to descend to and be ample on the coarse moduli space of the Deligne-Mumford stack delivered by [4, Theorem 1.14]. Neither the properness statement of Theorem 5.1 nor this ampleness statement is independently verified in the manuscript. If either has a gap, det(W) need not be big and the entire construction culminating in Theorem 5.10 collapses. The authors should provide a proof or a precise reference that covers this exact class and this exact descent statement.","section":"Theorem 5.1 and Proposition 5.3"},{"comment":"The proof applies Proposition 3.18 to the subsheaf K ⊂ eH^w on U to conclude that −c1(K) is pseudo-effective, and then uses det(j_*(K^∨)) to produce a morphism into Ω^{⊗n0 r_K}_S(log D). But Proposition 3.18 is stated for a coherent subsheaf of eH^k on a projective eS, while in Theorem 6.1 the sheaves eH^w and K are only defined on U = S \\ Z with codim_S Z ≥ 2, and their extension to S is not discussed. The later use of determinants and double duals requires K to extend to a coherent subsheaf of the reflexive extension of eH^w, and pseudo-effectivity of −c1(K) must be checked on S. This is likely fixable by a standard Hartogs-type extension argument, but as written it is a gap in the proof of the main theorem.","section":"Theorem 6.1, proof"}],"minor_comments":[{"comment":"In the statement of Theorem 4.1(2), the first occurrence of logarithmic poles says 'along D_g' while the final inequality is with Ω_V(log(D_f ∩ V)); the intended divisor is D_f, not D_g.","section":"Theorem 4.1"},{"comment":"The notation S'_o appears in 'V^n_{g_o,BM} on S'_o' but is never defined; it should presumably be the open set U or S \\ D_g.","section":"Section 4.2"},{"comment":"In the descent from \\tilde S to S, the display uses a mixture of σ_* and σ^* without making the base space of each sheaf explicit; in particular, 'σ_*O_{\\tilde S}⊗M^{-1} ⊂ O_S^{⊕N'}' and the subsequent line would be much clearer if the pushforwards and pullbacks were written with their domains and targets.","section":"Theorem 5.10, Last step"},{"comment":"The abstract says 'every schematic generically finite covering of any closed substack' while the body defines hyperbolicity using generically finite morphisms from smooth quasi-projective varieties; the terminology should be reconciled.","section":"Introduction and Definition 1.1"},{"comment":"There are several typos and small notational slips, e.g., 'natual' in Theorem 2.13(3), 'discription' before Theorem 3.16, 'to to' in the proof of Lemma 5.5, and the undefined 'S'_o' mentioned above; a careful proofreading pass is needed.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The main theorems are explicitly conditional on [4, Theorem 1.14] and on the ampleness statement of Proposition 5.3, both of which go beyond the published KSBA projectivity results. Since [4] is a preprint, I recommend that the editor ask the authors to confirm the precise status of these results and to provide a proof or a citation covering the exact class of (d, Φ_c, Γ, σ)-stable minimal models and the stacky descent of λ_{a,r} before final acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This paper is worth your attention if you work on hyperbolicity of moduli. It does something genuinely new: it builds Viehweg-Zuo Higgs sheaves using Borel-Moore cohomology of semistable families of pairs, and uses that machinery to prove stratified hyperbolicity for Birkar's moduli stack of stable minimal models. The main theorems are substantial—if correct, they answer a large part of Birkar's question and extend the Viehweg hyperbolicity conjecture from smooth families to all strata of KSBA compactifications. The Borel-Moore VMHS construction and the geometric description of its lower canonical extension are real contributions, and the overall architecture follows the standard Viehweg-Zuo/Campana-Paun strategy with new input at the right places.\n\nThe weak spots are real but mostly the kind that revision can fix. The paper depends critically on two black boxes: Birkar's properness theorem for the moduli stack and Proposition 5.3, the ampleness of the Kollár-type polarization λ_{a,r}. The proof of Theorem 1.1 needs det(W) to be big, which is exactly (ξ∘σ)^*λ_{a,r}. If either Birkar's moduli theorem or that ampleness statement has a hidden gap, the whole Viehweg-Zuo construction in Theorem 5.10 loses its entry point. That is not an internal contradiction, but it is load-bearing, and the paper does not independently verify these inputs. I also see smaller issues: Definition 2.1 permits Q-divisor coefficients while Lemma 2.8 and equation (3.3) apply sheaf duality as though D were integral Cartier; there are apparent sigma_*/sigma^* typos in Theorem 5.10 near (5.16); and the existence of the birationally admissible stratification is only sketched via Lemma 6.6 plus recursion, with strict admissibility defined only for smooth bases while strata are not required to be smooth. None of these by itself sinks the paper, but they need to be addressed.\n\nI would send this to a serious referee. The claims are important, the construction is original, and the flaws are addressable—assuming Birkar's moduli results hold up. If you work in the area, read it now; just don't cite the main theorem as established until the dependency on those black boxes is sorted out.","headline":"Serious new Viehweg-Zuo construction via Borel-Moore cohomology; main theorem is significant but depends on unverified ampleness and properness results of Birkar.","tokens_in":47938,"tokens_out":2215,"would_cite":true,"duration_ms":24392,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14D22","14D07","14C30","14J10","14D23"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that every stratum of a birationally admissible stratification of the moduli stack of stable minimal models is hyperbolic, via Borel-Moore cohomology and a logarithmic Higgs sheaf.","keywords":["moduli of stable minimal models","KSBA moduli stack","stratified hyperbolicity","log general type","logarithmic Higgs sheaf","Borel-Moore cohomology","variation of mixed Hodge structures","Deligne-Mumford stack"],"falsifier":"Construct a smooth quasi-projective base $S$ whose logarithmic canonical bundle $K_S+D$ is not big (for instance $\\mathbb P^1$ minus one or two points, or an elliptic curve minus a point) supporting a birationally admissible family of stable minimal models with generically finite classifying map to the moduli stack. Such a family would contradict the main theorem directly; alternatively, locate a closed substack of a stratum admitting a generically finite cover that is not of log general type.","tokens_in":46541,"feed_emoji":"📐","tokens_out":10109,"duration_ms":99785,"temperature":0.7,"pith_summary":"This paper proves that the moduli stack of stable minimal models — the compactification parameterizing semi-log-canonical pairs with semi-ample canonical bundle — is hyperbolic in a stratified sense. The main theorem says that any birationally admissible family of stable minimal models over a smooth quasi-projective base, whose classifying map to the moduli stack is generically finite, has base of log general type. Since a birationally admissible stratification always exists, each of its strata carries this hyperbolicity property: every generically finite cover of any closed substack of a stratum is of log general type. This extends the classical hyperbolicity program for families of polarized manifolds to the singular boundary strata of moduli of stable pairs, giving a partial answer to a question raised in the construction of these moduli spaces.","feed_headline":"Stable-model moduli: every admissible stratum is hyperbolic","feed_subtitle":"A Borel-Moore Higgs-sheaf argument proves every generically finite cover has log-general-type base.","key_machinery":"The central object is the Borel-Moore cohomology of a simple normal crossing pair $(X,D)$, $H^i_{BM}(X\\setminus D;\\mathbb{Q})$, studied through its family version as a graded-polarizable admissible variation of mixed Hodge structures. Its top Hodge piece is $O_X(K_X+D)$, so the determinant polarization $\\det f_*(O(rK_{X/S}+rB+raA))$ embeds into the lowest Hodge piece of the associated lower canonical logarithmic Higgs bundle. The Higgs field then propagates this subbundle through the Hodge filtration, giving the morphism to $\\Omega_S^{\\otimes m}(\\log D)$ that makes $K_S+D$ big. This is the classical Higgs-sheaf mechanism for hyperbolicity, rebuilt with Borel-Moore cohomology and mixed Hodge theory to handle singular fibers and boundary divisors.","core_discovery":"The central discovery is that the natural determinant line bundle on the moduli stack, $\\lambda_{a,r}=\\det f_*(r(K_{X/S}+B+aA))$, is not just a polarization but the seed of the hyperbolicity argument. The paper shows that, for a birationally admissible family, the Borel-Moore cohomology of a simple normal crossing log birational model carries an admissible variation of mixed Hodge structures whose lower canonical logarithmic Higgs bundle contains this determinant line bundle as its lowest Hodge piece. Iterating the Higgs field produces a non-zero morphism from a tensor power of the determinant bundle to a tensor power of the logarithmic cotangent bundle, twisted by a weakly positive sheaf. Because the determinant bundle is ample, its pullback is big, and the known positivity-to-bigness criterion forces $K_S+D$ to be big. The birationally admissible condition — existence of a simple normal crossing log birational model — is precisely what makes the Borel-Moore variation well-behaved enough to run this argument on every stratum, including the boundary.","pith_inferences":["If the same construction can be pushed past the birationally admissible condition, the failure is likely concentrated in degenerate fibers; the paper's own examples of a Lefschetz pencil over $\\mathbb P^1$ and an elliptic base with degenerating polarization suggest the hypothesis is close to optimal rather than merely technical.","The Borel-Moore Higgs bundle may carry more information than bigness: the same morphisms could yield non-vanishing of symmetric powers of the logarithmic cotangent sheaf, potentially implying Brody or algebraic hyperbolicity of the strata rather than only log-general-type hyperbolicity.","Because the stratification is chosen rather than canonical, one can test whether the hyperbolicity conclusion is independent of the choice of functorial desingularization; if different choices produce incompatible strata, the statement would have to be read as a property of the chosen filtration.","For other moduli stacks with an ample determinant-type line bundle and a well-behaved universal family, the same Borel-Moore mechanism would likely produce stratified hyperbolicity once the required Hodge-theoretic inputs are available."],"forward_implications":["The boundary strata of the moduli stack of stable curves are hyperbolic, including the deep boundary; in particular the whole stack $\\overline{\\mathcal M}_{g,n}$ is hyperbolic under this notion.","No birationally admissible family of stable minimal models with maximal variation can have base $\\mathbb P^1$, an abelian variety, or any other variety whose logarithmic canonical bundle is not big; such bases force isotriviality or non-admissibility.","The hyperbolicity statement covers principal and boundary strata uniformly, so the classical results for smooth canonically polarized families now extend to stable slc pairs with semi-ample canonical class.","Every generically finite cover of a closed substack of a stratum is of log general type, a strong stack-theoretic hyperbolicity that rules out large families of varieties mapping essentially finitely to the moduli stack."],"supporting_citations":[{"why":"Supplies the proper Deligne-Mumford stack and projective coarse moduli space of stable minimal models.","marker":"[4]"},{"why":"Semipositivity of direct images of pluricanonical bundles, used to prove the determinant line bundle is nef and ample.","marker":"[16]"},{"why":"Projectivity of the moduli space of stable log-varieties and ampleness of its polarization, a key external input for bigness.","marker":"[35]"},{"why":"Develops variation of mixed Hodge structure for simple normal crossing pairs and semipositivity of direct images, the technical backbone of the Borel-Moore construction.","marker":"[18]"},{"why":"Provides the criterion that a Higgs morphism from a big line bundle to a cotangent power forces $K_S+D$ to be big.","marker":"[6]"},{"why":"The original hyperbolicity argument for moduli of canonically polarized manifolds, whose structural outline the paper adapts.","marker":"[51]"},{"why":"Establishes hyperbolicity and isotriviality results over curves, including the degenerate-fiber examples that motivate the admissibility condition.","marker":"[49]"},{"why":"Gives the limiting mixed Hodge structure and monodromy weight filtration used to prove admissibility and construct the lower canonical extension.","marker":"[45]"},{"why":"Semi-positivity of kernels of Higgs bundles, used to show pseudoeffectivity of the relevant Chern classes.","marker":"[5]"}],"fun_headline_variants":["All admissible strata hyperbolic in stable-moduli","Hyperbolicity via Higgs sheaves on moduli strata","Stable-model stack: every stratum is hyperbolic","Determinant line bundle seeds hyperbolicity proof","Stratified hyperbolicity for stable minimal models"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof's starting point is the claim that the moduli stack is complete and carries an ample determinant line bundle, as asserted by the cited moduli theorem together with the cited semipositivity results; if that external foundation fails, the main theorem collapses.","fun_headline_variants_meta":{"raw":{"variants":["All admissible strata hyperbolic in stable-moduli","Hyperbolicity via Higgs sheaves on moduli strata","Stable-model stack: every stratum is hyperbolic","Determinant line bundle seeds hyperbolicity proof","Stratified hyperbolicity for stable minimal models"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000676,"raw_usage":{"total_tokens":3027,"prompt_tokens":849,"completion_tokens":2178,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":465,"completion_tokens_details":{"reasoning_tokens":2104}},"tokens_in":465,"tokens_out":2178,"duration_ms":16440,"temperature":1.0,"reasoning_tokens":2104,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T19:10:09.413345+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct a smooth quasi-projective base $S$ whose logarithmic canonical bundle $K_S+D$ is not big (for instance $\\mathbb P^1$ minus one or two points, or an elliptic curve minus a point) supporting a birationally admissible family of stable minimal models with generically finite classifying map to the moduli stack. Such a family would contradict the main theorem directly; alternatively, locate a closed substack of a stratum admitting a generically finite cover that is not of log general type.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Semipositivity of direct images of pluricanonical bundles, used to prove the determinant line bundle is nef and ample."},{"cited_title":"Fujino and T","cited_arxiv_id":null,"evidence_quote":"Develops variation of mixed Hodge structure for simple normal crossing pairs and semipositivity of direct images, the technical backbone of the Borel-Moore construction."},{"cited_title":"Campana and M","cited_arxiv_id":null,"evidence_quote":"Provides the criterion that a Higgs morphism from a big line bundle to a cotangent power forces $K_S+D$ to be big."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The original hyperbolicity argument for moduli of canonically polarized manifolds, whose structural outline the paper adapts."},{"cited_title":"Viehweg and K","cited_arxiv_id":null,"evidence_quote":"Establishes hyperbolicity and isotriviality results over curves, including the degenerate-fiber examples that motivate the admissibility condition."},{"cited_title":"Steenbrink, Limits of Hodge structures , Invent","cited_arxiv_id":null,"evidence_quote":"Gives the limiting mixed Hodge structure and monodromy weight filtration used to prove admissibility and construct the lower canonical extension."}],"review_version":2}