{"id":"46a962f2-0c12-4f46-8820-b5e5e7bd1c33","arxiv_id":"2507.07937","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"high","formal_verification":"none","parameter_count":3,"one_line_summary":"A moduli-theoretic framework for PDEs via D-Hilbert schemes and Spencer stability is introduced, but the advertised refinement of Donaldson-Uhlenbeck-Yau is a restatement of the classical result.","lead":"The authors construct moduli spaces of involutive systems of partial differential equations as geometric objects, introducing D-Hilbert and D-Quot functors and a notion of Spencer stability. They claim this framework refines the Donaldson-Uhlenbeck-Yau correspondence, but the proof explicitly reduces that refinement to the classical theorem itself.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Prop. 6.5's identification of Spencer slope with μ(E) is the critical failure: for a flat bundle, H^0(Sp(I∇)) is a local-system cohomology group, not a coherent sheaf, so 'deg' is undefined or zero, and Spencer stability does not reduce to ordinary slope stability. Theorem 6.2 depends on this.","rationale":"The paper's central advertised result (Theorem 6.2) depends on Proposition 6.5, which identifies the newly defined Spencer slope with the classical slope of the underlying flat bundle. The proof of this proposition contains the unsupported claim that deg(ker d∇) is determined by rank(E) and deg(E) via HRR. This is the weakest point: for a flat bundle, ker d∇ is a local system (or its cohomology), not a coherent sheaf, so its degree is not a well-defined input for the usual HRR formula; and in any formal interpretation its Chern classes vanish, so the claimed proportionality forces all Spencer slopes to be zero, which would make the stability inequality vacuous and break the equivalence with HYM existence. This is a genuine correctness risk, not merely a disagreement with consensus: the internal logic of §6.2.1 uses the numerical identification to translate Spencer-stability inequalities into μ(F) ≤ μ(E), and without it the proof is just an appeal to Donaldson–Uhlenbeck–Yau. The representability theorem 5.1 is also only sketched, but the slope issue is the most load-bearing because it invalidates the advertised application. The proposed concrete test—computing μSp for a flat extension on an elliptic curve—would settle whether the identification holds. Consequently, the reader's REJECT verdict stands.","tokens_in":58580,"tokens_out":14072,"duration_ms":177739,"concrete_test":"Take X an elliptic curve and E = O ⊕ O with the flat connection ∇ = d + A, where A = [[0, η], [0, 0]] for a nonzero holomorphic 1-form η. Compute the Spencer slope (4.7) directly: H^0(Sp(I∇)) = ker(d + [A, ·]) on O^4 equals C^2 (constant matrices [[a, b], [0, a]]), so μSp(I∇) = 0 if degree is read as 0 on constants. Repeat for the ∇-invariant sub-ideal corresponding to the O-subbundle: μSp(J) = 0. Thus the strict inequality in Spencer stability fails for a semistable flat bundle, contradicting the claimed reduction to μ(F) ≤ μ(E). If the paper's HRR identification were correct, the two slopes would differ by the formula; they do not. Also check the Atiyah class: for a flat bundle it vanishes, so any HRR-type formula expressing deg(ker d∇) as a linear function of deg(E) forces the constant to be 0, making all Spencer slopes equal.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Proposition 6.5 (§6.2.1) asserts that 'deg(ker d∇) is determined by rank(E) and deg(E) by Hirzebruch–Riemann–Roch', and uses this to replace the Spencer slope μSp(I∇) by the usual slope μ(E). This is the single step that makes Theorem 6.2 a statement about flat bundles rather than about the new Spencer formalism. The step is not justified and appears false. For a holomorphic flat bundle, ker d∇ is the sheaf of flat sections of End(E) — a local system; as an OX-module it is not coherent, so its 'degree' is not defined. If one formally vector-bundle-ifies it, the resulting flat bundle has vanishing Chern classes, hence degree 0, independently of deg(E). For E = O^r with the trivial connection on any X, ker d∇ = C^{r^2} (constant matrices), so μSp(I∇) = 0; for every ∇-invariant sub-ideal the same computation gives slope 0, making the strict Spencer-stability inequality vacuous. Thus the claimed equivalence with Gieseker polystability of the associated Higgs bundle cannot be derived from the Spencer slope as defined in (4.7). The proof of Proposition 6.5 also only tests ∇-invariant subbundles, whereas the Higgs-bundle stability condition used in the invoked DUY/Simpson correspondence requires θ-invariant subbundles; the paper does not establish that these coincide. Without Proposition 6.5, Theorem 6.2 is unsupported: it is simply an appeal to Donaldson–Uhlenbeck–Yau/Simpson rather than a refinement.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a moduli-theoretic framework for systems of PDEs in algebraic geometry. It defines D-ideal sheaves in infinite jet bundles, introduces Spencer regularity, D-Hilbert polynomials, and the notions of Spencer (semi)stability and Spencer slope, and claims that the corresponding D-Hilbert functor is representable by a finite-type ind-scheme (Theorem 5.1). The main application is Theorem 6.2: for a holomorphic flat bundle (E, ∇) on a compact Kähler manifold, the D-ideal I∇ encoding the flat-connection equation is Spencer-polystable if and only if E admits a Hermitian–Yang–Mills metric, which the authors present as a refinement of the Donaldson–Uhlenbeck–Yau correspondence.","tokens_in":59005,"tokens_out":4694,"duration_ms":53360,"significance":"If the main claims were established, the paper would introduce a genuinely novel formalism connecting the formal theory of PDEs with moduli problems and stability conditions, with potential applications to flat connections, Higgs bundles, and possibly K-stability. The authors engage seriously with the Spencer–Goldschmidt–Vinogradov tradition and propose concrete moduli functors with numerical invariants. However, the advertised results are not proven as they stand: the identification of Spencer slope with ordinary slope in the flat-connection example is unjustified, the proof of Theorem 6.2 invokes the classical Donaldson–Uhlenbeck–Yau and Simpson theorems rather than deriving a refinement, and the representability theorem rests on several proof sketches and omitted checks. The paper contains no machine-checked proofs, reproducible code, or parameter-free derivation; its central numerical comparison is exactly the point that needs independent verification and is not supplied.","major_comments":[{"comment":"The proof identifies the Spencer slope μSp(I∇), defined in (4.7), with the ordinary slope μ(E) by asserting that deg(ker d∇) is determined by rank(E) and deg(E) via Hirzebruch–Riemann–Roch. For a holomorphic flat bundle, ker d∇ is the local system of flat sections of End(E), not a coherent OX-module, so its degree is not defined in the usual sense; if the local system is converted into a vector bundle with flat connection, its Chern classes vanish and the degree is 0 independently of deg(E). For the trivial connection on O^r, for example, H^0(Sp(I∇)) consists of constant matrices and the Spencer slope is 0, making the strict Spencer-stability inequality vacuous. Since this identification is the only step connecting Spencer stability to Gieseker stability of the associated Higgs bundle, Theorem 6.2 is not established.","section":"§6.2.1, Proposition 6.5"},{"comment":"The comparison with Gieseker stability tests only ∇-invariant subbundles F with induced subconnections. The stability condition appearing in the Donaldson–Uhlenbeck–Yau/Simpson correspondence is Higgs stability of (E, θ), which requires θ-invariant subbundles. The paper does not prove that ∇-invariant subbundles coincide with θ-invariant subbundles, and consequently the Spencer-stability inequalities are not shown to be equivalent to the Higgs-slope inequalities used in the invoked correspondence.","section":"§6.2.1, proof of Proposition 6.5"},{"comment":"The claimed refinement of the DUY correspondence is not derived from the Spencer formalism: the proof assumes the classical Donaldson–Uhlenbeck–Yau theorem and Simpson's non-abelian Hodge theory to pass from Spencer-polystability to a harmonic metric and conversely. This makes the theorem, as stated, a reformulation of known results rather than a refinement of them. In addition, Proposition 6.6, which is used to glue stable summands, is stated without proof.","section":"§6.2.1, proof of Theorem 6.2"},{"comment":"The representability theorem rests on the uniform boundedness statement Proposition 5.3, whose proof cites Malgrange and Sweeney for the key bound, and on Proposition 5.4, whose proof contains the sentence 'It is a long check so we omit the details in full.' Proposition 3.14 is likewise justified by 'standard homological arguments.' Since the finite-type ind-scheme structure of Hilb^P_DX(J^∞_X E) is the first main result of the paper, these gaps leave the representability claim only partially documented.","section":"§5.1–5.2, Theorem 5.1"}],"minor_comments":[{"comment":"There are several typographical errors: 'Pesudogroups' in §1.2, 'conserve' for 'converse' in the introduction's statement of Theorem 6.2, 'unubstructedness' in §6.2.1, and 'Casetlnuovo-Mumford' before Definition 2.33.","section":"§1.2, §6.2.1, §2.5"},{"comment":"The quantity rank(I) is used in (4.6) before its precise definition as the leading functional rank is given in Proposition 4.23; the definition should be moved before (4.6) to avoid ambiguity.","section":"§4.3, equation (4.6)"},{"comment":"The references [NS] and [Ni] list the first author as 'Nitin, N.'; the correct name is Nitsure.","section":"References [NS], [Ni]"},{"comment":"Definition 4.22 compares the reduced D-Hilbert polynomials P̄_D(J) and P̄_D(I) without explicitly stating that the comparison is required for all sufficiently large n; this should be stated to make the definition unambiguous.","section":"§4.3, Definition 4.22"}],"recommendation":"reject","confidential_remarks":"The manuscript reads more like a research announcement than a complete proof of its main theorems. The proof of Theorem 6.2 appears to rely on the exact classical theorems it claims to refine, and Proposition 6.5 contains a questionable identification of Spencer slope with ordinary slope for flat bundles; this is load-bearing for the paper's central application. The representability theorem also depends on several results that are only sketched or invoked from the literature. I would not recommend sending the paper for revision until the slope comparison is corrected or replaced, the θ-invariance issue is resolved, and the omitted checks in Propositions 3.14, 5.4, and 6.6 are supplied."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The reader's take is close to mine, and the stress-test note lands on the load-bearing point. What is genuinely new here: the D-Hilbert and D-Quot functors, Spencer regularity, and Spencer (semi)stability (Definitions 4.22–4.24, (4.7)) are real definitions that I have not seen in this combination. The paper also does a serious job assembling D-geometry, jets, and Spencer cohomology into one formalism, and Theorem 5.13's tangent-space computation is a plausible starting point. I would not dismiss the program.\n\nThe problem is the advertised application. Proposition 6.5 identifies the Spencer slope of the flat-connection ideal with the usual slope of E using deg(ker d∇), with 'determined by rank(E) and deg(E) via Hirzebruch–Riemann–Roch'. That is not right. For a flat bundle, ker d∇ is the sheaf of flat sections of End(E), a local system, not a coherent OX-module; its degree is not defined. If you replace it by the associated flat vector bundle, the Chern classes vanish, so the degree is 0 regardless of E. On X = any base with E = O^r and the trivial connection, ker d∇ is the constant sheaf, so every Spencer slope is 0 and the stability inequalities are vacuous. The proof also only checks ∇-invariant subbundles, not the θ-invariant subbundles required by the Higgs-bundle stability in the DUY/Simpson correspondence it invokes. So Theorem 6.2, as stated, is not a refinement of DUY; it is an appeal to DUY plus an unsupported change of variables.\n\nThe paper is honest about some of its own gaps: Proposition 5.4 says 'long check ... omitted', Proposition 3.14 is relegated to 'standard homological arguments', and Proposition 6.6 is stated without proof. Those are proportionate warnings, but they mean Theorem 5.1 is a program sketch rather than a complete representability theorem. The boundedness arguments lean heavily on Malgrange and Sweeney; that may be fillable, but it is not written.\n\nWho should read it: people working on D-module moduli or the formal theory of PDEs will find the definitions and the organizing framework useful. Anyone citing Theorem 6.2 as evidence for Spencer stability should wait. If I were the editor, I would send this to a serious referee—the new objects deserve expert scrutiny—but I would expect either a major revision that fixes §6.2.1 or a split that presents the D-Hilbert program without the false application.","headline":"New D-Hilbert/Spencer-stability formalism, but the advertised flat-connection application is undercut by a false slope identification and the representability theorem is only sketched.","tokens_in":59559,"tokens_out":4868,"would_cite":false,"duration_ms":56885,"reading_group":"no","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14C05","14D20","35A30","53C07"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that the differential ideal of a flat connection is Spencer-polystable precisely when the bundle admits a Hermitian–Yang–Mills metric, and that involutive PDE ideals with fixed D-Hilbert polynomial form a representable…","keywords":["D-Hilbert scheme","Spencer stability","involutive PDE systems","Hermitian-Yang-Mills metric","flat connections","moduli spaces","formal integrability","D-modules"],"falsifier":"Take a rank-two flat bundle with nontrivial monodromy on a compact Riemann surface, compute the Spencer slope $\\mu_{\\mathrm{Sp}}(I_\\nabla)$ from $\\ker(d_\\nabla)$ inside $\\mathrm{End}(E)$, and compare it with the ordinary slope $\\mu(E)$; if the two slopes differ, or if a Spencer-stable $I_\\nabla$ corresponds to a destabilized bundle, the claimed equivalence fails.","tokens_in":58372,"feed_emoji":"📐","tokens_out":7487,"duration_ms":88259,"temperature":0.7,"pith_summary":"This paper tries to build a moduli space for partial differential equations themselves, treating a system of PDEs as an ideal sheaf in the infinite jet algebra and asking when such ideals can be classified by a scheme. The authors introduce the D-Hilbert and D-Quot functors for formally integrable, involutive ideal sheaves with a fixed D-Hilbert polynomial, and claim these are representable by a finite-type ind-scheme. The engine of the classification is Spencer (semi)stability, a slope condition defined through Spencer cohomology of the equation's symbol. As the main application, they claim that the differential ideal encoding a flat connection on a compact Kähler manifold is Spencer-polystable exactly when the underlying bundle admits a Hermitian–Yang–Mills metric, giving a PDE-theoretic restatement of the Donaldson–Uhlenbeck–Yau correspondence. If true, this puts flat-bundle stability and canonical-metric existence under one cohomological umbrella.","feed_headline":"Flat connections admit Hermitian–Yang–Mills metrics exactly when Spencer-stable","feed_subtitle":"A new moduli theory for PDE ideals puts the Donaldson–Uhlenbeck–Yau correspondence inside Spencer stability.","key_machinery":"The load-bearing object is the Spencer $\\delta$-complex attached to the symbol of a D-ideal, together with the resulting Spencer regularity degree $\\mathrm{reg}_{\\mathrm{Sp}}(I)$ and the Spencer slope $\\mu_{\\mathrm{Sp}}(I) = \\deg H^0(\\mathrm{Sp}(I))/\\mathrm{rank}(I)$. The D-Hilbert polynomial packages the Spencer cohomology of the symbol as a numerical invariant, and Spencer (semi)stability compares reduced D-Hilbert polynomials of involutive subideals. Involutivity and formal integrability ensure that the D-Hilbert polynomial is polynomial and that Spencer regularity is bounded across families, which is what carries the representability proof.","core_discovery":"The central claim is that the formal theory of PDEs can be organised into a moduli problem whose numerical invariant is the D-Hilbert polynomial and whose stability condition is Spencer stability. For the flat-connection equation (curvature zero), the associated D-ideal $I_\\nabla$ has a Spencer complex whose zeroth cohomology is $\\ker(d_\\nabla \\colon \\mathrm{End}(E) \\to \\Omega^1 \\mathrm{End}(E))$, and the paper argues that Spencer stability of $I_\\nabla$ matches Gieseker stability of the Higgs bundle attached to $\\nabla$ by a harmonic metric, with Spencer slopes comparing through the ordinary slopes of invariant subbundles. Theorem 6.2 then states that $I_\\nabla$ is Spencer-polystable if and only if $E$ admits a Hermitian–Yang–Mills metric, while Theorem 5.1 states that the functor of formally integrable, involutive D-ideal sheaves with fixed D-Hilbert polynomial is representable by a finite-type ind-scheme with a compatible D-action.","pith_inferences":["A testable extension would be to compute $\\mu_{\\mathrm{Sp}}(I_\\nabla)$ directly for a flat bundle with nontrivial monodromy and compare it with the ordinary slope of $E$; if the two slopes do not agree, Theorem 6.2 would need a modified slope rather than the stated Hirzebruch–Riemann–Roch identification.","The D-Hilbert construction should be comparable numerically with known moduli of flat bundles: if its points correspond to sub-local systems of schemes, its connected components may match the coarse moduli of semistable flat bundles from non-abelian Hodge theory, giving a purely algebraic route to those moduli spaces.","The same Spencer-stability formalism could be applied to other geometric PDEs, such as the Monge–Ampère equation on a Fano manifold, replacing 'Hermitian–Yang–Mills metric exists' by 'Kähler–Einstein metric exists'; the paper states this as a conjecture, so any verification would be a genuinely new result.","Because Spencer regularity gives an effective bound on the order at which a differential ideal is determined, the moduli scheme may admit explicit charts at finite jet level, making the construction algorithmic in the spirit of Janet–Riquier theory."],"forward_implications":["If Theorem 5.1 is correct, the classical Hilbert scheme extends to differential-algebraic geometry: formally integrable, involutive PDE systems with fixed D-Hilbert polynomial form a finite-type ind-scheme with a compatible D-action.","If Theorem 6.2 is correct, the Donaldson–Uhlenbeck–Yau correspondence becomes a special case of Spencer stability, so flat-bundle moduli can be described by stability of the connection ideal rather than by slope stability of the bundle alone.","The boundedness result, uniform Spencer regularity across families, would give a flattening stratification for PDE ideals analogous to Mumford's boundedness, with recursive bounds depending only on dimension, rank, and order.","The tangent-space computation identifies first-order deformations of a D-ideal with $\\mathrm{Hom}_{D}(I, O(J^\\infty E)/I)$ and obstructions with truncated Spencer cohomology, supplying the deformation theory needed for a derived analogue of the moduli space."],"supporting_citations":[{"why":"Supplies the integrability criteria and formal theory of analytic nonlinear PDEs that underpin formal integrability of the D-ideals.","marker":"[G]"},{"why":"Introduces the Spencer complex and Spencer cohomology used to define Spencer regularity and Spencer stability.","marker":"[Sp]"},{"why":"Provides the classical Hilbert scheme and Quot-scheme representability results that the D-Hilbert functor generalizes.","marker":"[Gro3]"},{"why":"Gives the analytic correspondence between stable bundles and Yang–Mills connections that the paper refines.","marker":"[Do]"},{"why":"Gives the existence of Hermitian–Yang–Mills metrics on stable bundles, the classical statement being compared in Theorem 6.2.","marker":"[UY]"},{"why":"Supplies the moduli of flat bundles and non-abelian Hodge theory used to pass from flat connections to Higgs bundles in the proof.","marker":"[Si]"},{"why":"Establishes that Cartan involutivity equals Mumford regularity, the bridge for the Spencer-regularity criterion.","marker":"[Ma3]"},{"why":"Provides the Gieseker and slope stability formalism for sheaves that Spencer stability is modeled on.","marker":"[HL]"}],"fun_headline_variants":["Spencer stability refines Hermitian-Yang-Mills for flat connections","New moduli theory links PDE ideals to Spencer stability","Flat connections get Hermitian-Yang-Mills iff Spencer-polystable","Donaldson-Uhlenbeck-Yau correspondence refined by Spencer stability","Spencer stability unifies geometric PDEs and gauge theory"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The equivalence rests on identifying the Spencer slope of the flat-connection ideal with the ordinary slope of $E$, via the claim that $\\deg(\\ker d_\\nabla)$ is determined by $\\mathrm{rank}(E)$ and $\\deg(E)$ through Hirzebruch–Riemann–Roch; if this identification fails, the two sides of Theorem 6.2 no longer match.","fun_headline_variants_meta":{"raw":{"variants":["Spencer stability refines Hermitian-Yang-Mills for flat connections","New moduli theory links PDE ideals to Spencer stability","Flat connections get Hermitian-Yang-Mills iff Spencer-polystable","Donaldson-Uhlenbeck-Yau correspondence refined by Spencer stability","Spencer stability unifies geometric PDEs and gauge theory"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000845,"raw_usage":{"total_tokens":3679,"prompt_tokens":943,"completion_tokens":2736,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":559,"completion_tokens_details":{"reasoning_tokens":2647}},"tokens_in":559,"tokens_out":2736,"duration_ms":22577,"temperature":1.0,"reasoning_tokens":2647,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T18:30:25.578207+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a rank-two flat bundle with nontrivial monodromy on a compact Riemann surface, compute the Spencer slope $\\mu_{\\mathrm{Sp}}(I_\\nabla)$ from $\\ker(d_\\nabla)$ inside $\\mathrm{End}(E)$, and compare it with the ordinary slope $\\mu(E)$; if the two slopes differ, or if a Spencer-stable $I_\\nabla$ corresponds to a destabilized bundle, the claimed equivalence fails.","supporting_citations":[],"review_version":1}