{"id":"997bcf1b-5791-4267-b4ce-2a0dd320ea15","arxiv_id":"2411.16912","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The authors construct meromorphic Gieseker Higgs bundles to build flat and proper Hitchin fibrations over the moduli of stable pointed curves, including fixed nilpotent residue versions.","lead":"Using semistable modifications of pointed curves and Gieseker vector bundles, the authors extend the moduli spaces of meromorphic Higgs bundles to the compactification of the moduli of stable curves, with a flat and proper Hitchin map. This gives a boundary-complete version of the classical Hitchin integrable system, including the case of fixed nilpotent residues at the punctures.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem A rests on the unpublished [HLH]: Propositions 2.9/2.10 feed Lemma 2.17, Lemma 4.28, and hence strict monotonicity in Proposition 4.30; without them the good moduli space and flat proper Hitchin morphism are not established.","rationale":"I agree with the reader's assessment. The strongest claim is exactly Theorem A, and the most load-bearing assumption is the dependence on the unpublished [HLH] results, especially Propositions 2.9 and 2.10. These enter the proof at several critical junctures: Lemma 2.17 uses Proposition 2.9 to show that the pushforward to the stabilization is proper schematic; Lemma 4.28 uses Proposition 2.10 to prove eventual ampleness of the Cornalba line bundle; and Proposition 4.30 uses Lemma 4.28 to establish strict Theta- and S-monotonicity. Without strict monotonicity, Theorem 4.16 cannot produce the good moduli space, and without the good moduli space the flat proper Hitchin morphism in Theorem 4.37 is not established. I also examined a possible internal concern in Lemma 2.17, where the proof uses the Hom-scheme of a non-locally-free kernel; that step can likely be repaired by viewing the vanishing condition as a closed zero-locus condition on the base, so I do not regard it as fatal. I found no manufactured data, no circular reasoning within the paper itself, and no obvious inconsistency in the dimension computations. The paper has independent supporting ingredients: the published criteria of AHLH23 and HLHJ24, the published BDD22 construction of the modified Hitchin base, and a detailed BNR-type description over the open locus U. Nevertheless, the central theorem remains conditional on external unpublished inputs from a coauthor. Therefore the reader's CONDITIONAL verdict is the correct one, and my stress-test does not change it.","tokens_in":38140,"tokens_out":16076,"duration_ms":178279,"concrete_test":"Obtain a version of [HLH] and verify Proposition 2.10 verbatim: for every quasicompact scheme T equipped with a morphism T to Coh^N_{g,n}, the pullback of L^{Cor,m} to GBun^N_{g,n} \times_{Coh^N_{g,n}} T is T-ample for all sufficiently large m. Then check that the proof of Proposition 4.30 yields the strict inequality in Definition 4.20(M3) using only Lemma 4.28 and [HLH23, Prop. 3.6]; in particular, test the case where Y is the annular degeneration ST_R, the central curve is a genus-2 nodal curve with a chain of two rational components, and the Gieseker bundle has rank 2. If the amplitude or the monotonicity inequality fails, Theorem 4.37 does not follow; if it passes, the conditional verdict is justified.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim, Theorem A (Theorems 4.37 and 4.38), requires an open semistable locus admitting a good moduli space and a flat proper Hitchin morphism. The existence step applies Theorem 4.16, which needs strict Theta- and S-monotonicity (Proposition 4.30), HN boundedness (Proposition 4.35), and the valuative criterion for properness (Proposition 2.23). The least secure input is the unpublished companion manuscript [HLH] by one of the authors. Proposition 2.9 (smoothness of GBun^N and schematic properness of GBun^N to Coh^N) is used in Lemma 2.17 to prove that the pushforward to the stabilization is proper schematic, which is needed for Proposition 2.23. Proposition 2.10 (eventual T-ampleness of the Cornalba line bundle L^{Cor,m} on GBun^N \times_{Coh^N} T) is used in Lemma 4.28 to get the corresponding ampleness on GHiggs^N \times_{Higgstf} T; Lemma 4.28 is then the source of the epsilon-term in Proposition 4.30 and of the eventual positivity of LGies,m + epsilon LCor,m along the closure Sigma. If Proposition 2.10 fails, or if [HLH] is not available, the proof of strict monotonicity is incomplete and Theorem 4.16 does not produce the good moduli space; consequently H would not be shown flat and proper. The paper supplies no independent proof of these stack-theoretic statements. This is not an internal contradiction, but it is a load-bearing external dependency on unpublished work by a coauthor.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces the stack GHiggs^N_{g,n} of rank N meromorphic Gieseker Higgs bundles over prestable n-pointed curves and studies its geometry over the moduli stack Mbar_{g,n} of stable pointed curves. It defines a Hitchin morphism H to the log-canonical Hitchin base A, proves a BNR-type description of the fibers over the open locus U of allowable nodal spectral covers, and develops a residue theory leading to modified Hitchin bases A_{O•} for fixed nilpotent conjugacy classes, following [BDD22]. The main theorems, Theorems 4.37 and 4.38, assert that open semistable loci admit relative good moduli spaces GHiggs^{N,χ}_{g,n} and GHiggs^{O•,χ}_{g,n} over Mbar_{g,n}, and that the induced Hitchin morphisms to A and A_{O•} are flat and proper. The existence proof follows the Θ-stratification framework of Alper–Halpern-Leistner–Heinloth, using a numerical invariant μ built from Gieseker line bundles and Cornalba line bundles, with strict monotonicity, Harder–Narasimhan boundedness, and the valuative criterion for properness as the key ingredients.","tokens_in":38513,"tokens_out":14091,"duration_ms":152560,"significance":"If the main results are correct, this is a substantial contribution: it provides a universal flat and proper extension of the relative meromorphic Hitchin fibration over Mbar_{g,n}, including the nilpotent-residue symplectic leaves that are relevant to class S theories, and it gives a spectral-correspondence description over an explicit open locus. The paper contains many useful concrete computations: the relative Fuchs relations, dimensions of the residue-constrained stacks, Cohen–Macaulayness statements, and a detailed description of the Gieseker modifications in the spectral-cover picture. The main caveat is that the stack-theoretic foundations—smoothness and properness of the Gieseker bundle stack, and eventual ampleness of the Cornalba line bundles—are quoted from the unpublished companion manuscript [HLH] by one of the authors. Those inputs are load-bearing for the good moduli space construction. There is no indication of fitted parameters or circular reasoning in the numerical invariant itself, but the paper is not self-contained on its key foundational statements.","major_comments":[{"comment":"The central stack-theoretic inputs are quoted in full from the unpublished manuscript [HLH] by one of the authors. Proposition 2.9 is used in Lemma 2.17 to prove that the stabilization pushforward ϖ : GHiggs^N_{g,n} → Higgs^N_{g,n} is schematic and proper, and Proposition 2.10 is used in Lemma 4.28 to prove eventual ampleness of the Cornalba line bundle L^{Cor,m} on the relevant fiber products. Both statements feed directly into the strict monotonicity result (Proposition 4.30) and hence into the application of Theorem 4.16 that produces the good moduli spaces and the flat proper Hitchin morphisms in Theorem A. Because [HLH] is not available to the reader and is not proved or even summarized in this manuscript, the hypotheses of Theorem 4.16 are not verifiable from the paper alone. This is a load-bearing external dependency and should be addressed explicitly, for example by including the needed statements as proved appendices or by making [HLH] publicly available before publication.","section":"Section 2.2, Propositions 2.9 and 2.10"},{"comment":"The proof of strict Θ- and S-monotonicity is only a sketch and does not verify the key inequality. The argument after the construction of the closure Σ reduces the desired monotonicity to the positivity of the formal line bundle LGies,m + ε LCor,m on Σ, and then states that any two Gm-equivariant points p1, p2 of Σ_o yield graded points g1, g2 with the same norm b(g1) = b(g2). The justification given is that the underlying torsion-free Higgs sheaves are generically isomorphic on eC_o. However, the quadratic norm b is computed from the ranks of the graded pieces of the associated graded sheaf, and generic isomorphy of the ungraded underlying sheaves does not by itself determine those graded ranks. Since strict monotonicity is exactly the inequality between the values of μ at 0 and at ∞ for such graded points, this missing justification affects the application of Theorem 4.16 and therefore the existence of the good moduli spaces and the flat/proper Hitchin morphisms in Theorems 4.37 and 4.38.","section":"Section 4.3, Proposition 4.30"},{"comment":"The BNR-type statement for the open locus U is advertised as a main structural result, but its proof is abbreviated at a load-bearing point. After establishing properness, quasifiniteness is proven by identifying the fiber of c over a point of U with certain line bundles on 'allowed' semistable modifications and invoking [EP16, Thm. 6.1]. The reduction from Gieseker Higgs bundles to line bundles on such modifications is described informally and relies on a local étale computation over U; as written, the compatibility between the Gieseker vector bundle condition on the pushforward and the quasistable-modification description is asserted rather than fully proved. This does not affect the main flatness/properness theorems, but it is a significant secondary claim that would benefit from a complete proof or a precise reference.","section":"Section 2.5, Proposition 2.25"}],"minor_comments":[{"comment":"The displayed condition after 'with N1 ≥ N2 ≥' is incomplete; it should read N1 ≥ N2 ≥ ... ≥ Nl.","section":"Definition 3.7"},{"comment":"The example 'm − εm^2 > 0' appears inconsistent with the stated total order on R[m,ε]: for any fixed small h > 0, the polynomial m − h m^2 is eventually negative in m. Please correct the example or clarify the intended ordering.","section":"Notation 4.7"},{"comment":"The notation GHiggs_{g,n} is used without the superscript N in several places in the proof, which is confusing because the rank N is fixed throughout; please make the notation uniform.","section":"Proof of Proposition 2.25"},{"comment":"The equality of sections A_{O•}(T) ≅ A_{O•}^{nv}(T) is proved for flat T → Mbar_{g,n} by reducing to pushforwards over the universal curve; it would be helpful to state explicitly that the universal curve C → Mbar_{g,n} is flat, so the lemma applies to the universal family and to the base changes used in Section 3.3.","section":"Lemma 3.14"}],"recommendation":"major_revision","confidential_remarks":"The paper depends for its main theorem on two foundational inputs from [HLH], an unpublished companion manuscript by the second author. The dependency is unusually concentrated: Propositions 2.9 and 2.10 are exactly the smoothness, properness, and ampleness statements that make Lemma 2.17 and Lemma 4.28 work, and without them Theorem 4.16 cannot be applied. I would recommend that the editors require the authors either to include proofs of these statements in the present paper or to provide a public version of [HLH] before acceptance. The strict monotonicity proof in Proposition 4.30 is also currently too compressed to be checked independently."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper does something real: it defines meromorphic Gieseker Higgs bundles and uses them to construct flat proper Hitchin morphisms over the full Deligne–Mumford stack of stable pointed curves, including the fixed-nilpotent-residue version over the BDD bases. That is a genuine advance beyond the smooth-curve setting, the one-parameter degenerations in [BBN16], and the n=0 case in [BBDP23]. The BNR-type description of general fibers over the allowable nodal locus U via compactified Jacobians is a useful structural result, and the relative residue theorem is handled carefully. The normalization-gluing dimension computations for Higgs bundles on semistable curves are the kind of solid, checkable work that makes the paper feel trustworthy.\n\nThe soft spot is exactly where the reader put it. Propositions 2.9 and 2.10 are quoted from the unpublished [HLH], and they feed Lemma 2.17, Lemma 4.28, and ultimately the strict monotonicity proof in Proposition 4.30. The monotonicity argument is not really self-contained; it reduces to [HLH23, Prop. 3.6] after constructing the relevant closures. If [HLH] is correct, the chain holds. If it is not available, the existence of the good moduli space and the flat proper Hitchin morphism does not follow from this paper alone. That is a load-bearing dependency, not a minor gap. I do not see internal contradictions, fitted parameters, or suspicious data. The dimension and flatness arguments over the base look sound, and the use of miracle flatness is legitimate when source and target are Cohen–Macaulay/smooth. The fixed-residue flatness argument via the scaling action and equidimensional fibers is standard and appears correct.\n\nWho is this for? People working on moduli of Hitchin systems, class S theories, and compactified Jacobians. A serious referee should engage, but the referee must have access to [HLH] or the authors should include the needed statements with proofs as an appendix. I would not accept this paper as is; I would send it to review with a request that the companion manuscript be posted or the key results proved inline.","headline":"A serious, well-structured construction of the universal meromorphic Hitchin system over stable pointed curves; the central theorem is plausible but its proof leans on unpublished [HLH], so the paper deserves review but should not be accepted without those inputs being made available.","tokens_in":39038,"tokens_out":1626,"would_cite":true,"duration_ms":18092,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14H10","14H60","14D20","14D23","14H70"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper constructs a flat, proper extension of the meromorphic Hitchin fibration over the moduli stack of stable pointed curves, for all ranks and Euler characteristics, and the same for fixed nilpotent residue classes.","keywords":["meromorphic Higgs bundles","Gieseker vector bundles","Hitchin fibration","stable pointed curves","moduli stacks","compactified Jacobians","nilpotent residues","Theta-stratifications"],"falsifier":"A concrete check is the fiber-dimension formula: flatness of $H: GHiggs^{\\mathcal{O}_\\bullet,\\chi}_{g,n} \\to A_{\\mathcal{O}_\\bullet}$ forces every fiber to have dimension $N^2(g-1)+\\frac{1}{2}\\sum_i \\dim(O_i)$. Computing this dimension at a boundary point where the spectral cover is not nodal would settle the claim; a fiber of larger dimension, or a failure of the relative moduli space to be flat over $\\overline{M}_{g,n}$, would refute Theorem A.","tokens_in":37905,"feed_emoji":"","tokens_out":13045,"duration_ms":113743,"temperature":0.7,"pith_summary":"This paper proves that the relative moduli space of semistable meromorphic Higgs bundles over smooth pointed curves admits a flat extension over the compactified moduli stack of stable pointed curves, for every rank and Euler characteristic. The extended objects are meromorphic Gieseker Higgs bundles, and the usual Hitchin morphism extends to a flat and proper map, both for the full moduli space and for the closed symplectic leaves obtained by fixing nilpotent conjugacy classes of the residues. Over the open locus where the spectral cover is nodal and etale over the nodes, the fibers are described by compactified Jacobians, giving a boundary analogue of the BNR correspondence. The paper therefore extends the meromorphic Hitchin integrable system to degenerations of the underlying curve.","feed_headline":"Hitchin fibration extends flatly over all stable pointed curves","feed_subtitle":"New moduli spaces carry the meromorphic Hitchin map over degenerating curves, with fibers as compactified Jacobians.","key_machinery":"The central object is the stack $GHiggs^N_{g,n}$ of meromorphic Gieseker Higgs bundles: a semistable $n$-pointed curve together with a vector bundle satisfying Gieseker's three conditions (surjective counit, $\\varphi$-ample determinant, torsion-free pushforward to the stabilization) and a Higgs field $\\psi: E \\to E \\otimes \\omega^{\\mathrm{log}}$. The proof machinery is the numerical invariant $\\mu = (-\\mathrm{wt}(L_{\\mathrm{Gies},m}) - \\epsilon\\,\\mathrm{wt}(L_{\\mathrm{Cor},m}))/\\sqrt{b}$ built from the Gieseker and Cornalba line bundles, for which the paper proves strict $\\Theta$- and $S$-monotonicity and Harder-Narasimhan boundedness; this yields the good moduli spaces and the flat proper Hitchin morphisms. The spectral cover $S \\to C$ carries the fiber description over the allowable locus $U$, where the fibers become compactified Jacobians.","core_discovery":"The paper's central claim, stated as Theorem A (Theorems 4.37 and 4.38), is that for every rank $N$ and Euler characteristic $\\chi$ there is a moduli-theoretic flat family $GHiggs^{N,\\chi}_{g,n} \\to \\overline{M}_{g,n}$ that restricts over the smooth locus to the usual relative Dolbeault moduli space of semistable meromorphic Higgs bundles, together with an extension of the Hitchin morphism $H: GHiggs^{N,\\chi}_{g,n} \\to A$ that is flat and proper. The same statement holds when one fixes an $n$-tuple $\\mathcal{O}_\\bullet$ of nilpotent conjugacy classes for the residues, with the Hitchin base replaced by the vector-bundle degeneration $A_{\\mathcal{O}_\\bullet}$ of the naive base. Over the open locus $U$ of data whose spectral cover is nodal and etale over the nodes, the fiber of $H$ is shown to be a compactified Jacobian of the spectral cover, so the ordinary BNR description of fibers persists at the boundary.","pith_inferences":["Beyond the paper's own claims, the stack-level flatness proofs in Section 3 do not use nilpotency, so the same construction should extend to non-nilpotent conjugacy classes once the Hitchin base is twisted appropriately; the authors note this as an open problem.","Beyond the paper's own claims, the eventual log-Poisson and log-symplectic forms promised in the sequel would turn the flat family of compactified Jacobians over the allowable locus into a degeneration of the classical integrable system itself, not merely of its moduli space.","Beyond the paper's own claims, a testable extension is to compute the monodromy of the compactified-Jacobian fibration around the non-allowable boundary of the Hitchin base; the BNR description should fail precisely where the spectral cover acquires non-etale nodes."],"forward_implications":["The relative moduli space of semistable meromorphic Higgs bundles, previously defined only over smooth pointed curves, now has a flat extension over the whole moduli stack of stable pointed curves.","The Hitchin morphism extends to a flat and proper map from this extension to the universal Hitchin base, so the meromorphic Hitchin fibration degenerates in a controlled way.","For each fixed nilpotent conjugacy class of residues, the corresponding closed symplectic leaf also extends to a flat and proper family over the moduli stack, with the twisted Hitchin base as target.","Over the allowable nodal spectral locus, the fibers of the Hitchin morphism are compactified Jacobians of the spectral cover, giving a boundary analogue of the BNR correspondence.","The Harder-Narasimhan Theta-stratification of the stack of meromorphic Gieseker Higgs bundles shows that the semistable loci are of finite type and admit separated good moduli spaces over the moduli stack."],"supporting_citations":[{"why":"It supplies the unpublished smoothness, properness, and eventual ampleness results for the Gieseker bundle stack that underlie Lemmas 2.17 and 4.28.","marker":"[HLH]"},{"why":"It provides the numerical-invariant and Theta-stratification criteria used to construct the good moduli spaces.","marker":"[HL22]"},{"why":"It supplies the semistable reduction and properness criteria that turn the Hitchin morphism into a proper map.","marker":"[AHLH23]"},{"why":"It supplies the infinite-dimensional GIT strategy, the affine-Grassmannian presentation, and the Gieseker line-bundle framework on the stack of sheaves.","marker":"[HLHJ24]"},{"why":"It constructs the twisted Hitchin base as a vector bundle over the moduli stack, the target of the nilpotent-residue Hitchin morphism.","marker":"[BDD22]"},{"why":"It is the classical spectral-correspondence result whose boundary analogue is established over the allowable locus.","marker":"[BNR89]"},{"why":"It provides the uniqueness of quasistable modifications and the compactified-Jacobian description used in the fiber analysis over the allowable locus.","marker":"[EP16]"},{"why":"It introduces the Gieseker-Higgs stack in the unpunctured case and supplies the spectral cover factorization used in proving properness of the Hitchin morphism.","marker":"[BBDP23]"}],"fun_headline_variants":["Compactified Hitchin fibration over stable curves, flatly","New moduli spaces with compactified Jacobian fibers","Hitchin morphism flat proper on universal compactification","BNR correspondence extended to nodal spectral covers","Meromorphic Higgs moduli compactified with flat Hitchin map"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the unpublished companion manuscript [HLH] correctly proves smoothness, schematic properness, and eventual ampleness for the stack of Gieseker vector bundles; if that premise gives way, the existence of the good moduli spaces and of the flat proper Hitchin morphisms in Theorem A does not follow.","fun_headline_variants_meta":{"raw":{"variants":["Compactified Hitchin fibration over stable curves, flatly","New moduli spaces with compactified Jacobian fibers","Hitchin morphism flat proper on universal compactification","BNR correspondence extended to nodal spectral covers","Meromorphic Higgs moduli compactified with flat Hitchin map"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000239,"raw_usage":{"total_tokens":1504,"prompt_tokens":922,"completion_tokens":582,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":538,"completion_tokens_details":{"reasoning_tokens":503}},"tokens_in":538,"tokens_out":582,"duration_ms":6863,"temperature":1.0,"reasoning_tokens":503,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T12:45:02.794075+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete check is the fiber-dimension formula: flatness of $H: GHiggs^{\\mathcal{O}_\\bullet,\\chi}_{g,n} \\to A_{\\mathcal{O}_\\bullet}$ forces every fiber to have dimension $N^2(g-1)+\\frac{1}{2}\\sum_i \\dim(O_i)$. Computing this dimension at a boundary point where the spectral cover is not nodal would settle the claim; a fiber of larger dimension, or a failure of the relative moduli space to be flat over $\\overline{M}_{g,n}$, would refute Theorem A.","supporting_citations":[],"review_version":1}