{"id":"8e156305-be55-4d41-8c2b-9a9ec676e9e8","arxiv_id":"2607.22074","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For rank-r connections with pole order at most 2, nilpotent leading term, and non-resonant semisimple monodromy, every connected component of the coarse moduli space is the affine space n⊕n, of dimension r(r−1).","lead":"This paper proves that a class of differential equations with mild singularities — meromorphic connections with a pole of order two and nilpotent leading part — has a moduli space that is exactly a complex affine space, with an explicit normal form. It is a structural first step toward a proposed 'atom' birational invariant of complex projective varieties built from quantum connections.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 4.1 appears internally sound; the load-bearing unproven step is identifying NSQC_c objects with actual Dubrovin-connection atoms, which the paper itself defers.","rationale":"The reader's weakest assumption identifies exactly the same burden. The internal algebra of Theorem 4.1 appears to hold together: Lemma 4.4's cokernel computation is correct; the z^{-2} coefficient is not in the image of L_μ, so α_ji is invariant; the non-resonance c_i≠c_j gives a regular semisimple monodromy, making the Z^r parametrization reasonable. The main obstruction to the announced result is external: the paper does not prove that Dubrovin connections of smooth projective varieties decompose into NSQC_c-objects. The §1 discussion invokes stationary phase and exponential-type Hodge structures, but the equivalence relation for atoms is explicitly left open. Non-resonance is assumed, never verified for the motivating Fano-type examples. This does not falsify Theorem 4.1; it limits the scope. Because the paper itself acknowledges the limitation, a CONDITIONAL verdict is appropriate; no change from the reader's verdict.","tokens_in":13867,"tokens_out":44779,"duration_ms":452254,"concrete_test":"Compute the formal Turrittin–Hukuhara–Levelt decomposition of the Dubrovin connection of a concrete non-semisimple Fano threefold (e.g., one of the Fano 3-folds with h^{1,2}>0 in the Graded Ring Database, using its known Picard–Fuchs/Gromov–Witten data). Check: (1) the decomposition is ⊕_a (d + a/z^2 dz)⊗∇^a_reg with ∇^a_reg regular singular; (2) after subtracting the simple part of K_{−2}, the summands satisfy (Cond1)–(Cond3) with the flag given by the monodromy filtration; (3) the monodromy eigenvalues c_i are pairwise distinct and none equal 1. If any check fails for a representative example, the theorem's announced application to atoms is unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The strongest claim is Theorem 4.1, but the paper's announced subject is the moduli of atoms of complex projective varieties. The proof of Theorem 4.1 is a derivation about the abstract groupoid NSQC_c; after spot-checking the coker computation (Lemma 4.4), the invariance of the z^{-2} coefficient, and the filtration argument in Lemma 4.2, I do not find a fatal internal error. (The displayed formula for ℓ_ii in §4 appears to have a typo — the intended gauge is exp(∫(μ_i/ζ − d_ii)dζ) — and the r≥3 part of Prop 2.1 is handed to the reader, but these are repairable.) The load-bearing unproven premise is the §1 reduction: that actual quantum connections of smooth projective varieties satisfy (Cond1)–(Cond3) and the non-resonance condition c_i≠c_j. The stationary-phase decomposition (1) is asserted heuristically, the 'up to simple endomorphisms' reduction of K_{−2} is not proved, and the flag condition is called a mere rigidification. Section 1 itself states that the atom equivalence relation 'is yet to be explored.' Thus if the motivating examples fail (Cond1)–(Cond3) or are resonant, Theorem 4.1 remains a theorem about an abstract groupoid but does not deliver 'moduli of atoms of varieties.'","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper introduces a groupoid NSQC_c of framed meromorphic connections on a rank-r holomorphic vector bundle over a disc: regular singularity at 0 (Cond1), pole order at most two (Cond2), and a basis at z=1 whose flag is preserved by monodromy with prescribed eigenvalues c_i (Cond3). The main theorem (Thm 4.1) states that if c_i ≠ c_j and c_i ≠ 1, the connected components of the coarse moduli space of NSQC_c are isomorphic to n ⊕ n, an affine space of dimension r(r−1). The proof chain is: Prop 3.1–3.2 (normal form for regular-singular connections with an invariant flag, yielding D_n ∈ n for n < −1 and D_n ∈ b for n ≥ −1), reduction to polynomial connections in §4, and Prop 4.3 with Lemma 4.4 (the operator L_μ puts each orbit into the unique normal form α_{ji} z^{−2} + β_{ji} z^{m(j,i)−1}). The introduction motivates the conditions via 'atoms' of Dubrovin connections of smooth projective varieties and sketches a birational-invariant formalism, while explicitly deferring the definition of the atom equivalence relation.","tokens_in":14067,"tokens_out":36049,"duration_ms":318151,"significance":"The normal-form theorem itself is valuable and, as far as the reviewer checked, internally sound: the L_μ action (19), the coker computation in Lemma 4.4, and the invariance of the z^{−2} coefficient in Prop 4.3 are explicit and verifiable; the derivation is self-contained given Deligne's Riemann–Hilbert correspondence, Fuchs' criterion, and Levelt's normal form, and no parameter is fitted to an external target. The paper provides canonical affine coordinates (α, β) for a natural class of non-semisimple connections, a concrete step beyond the semisimple theory. However, the significance for the announced subject — moduli of atoms of complex projective varieties — is conditional on the §1 reduction from quantum connections to (Cond1)–(Cond3) and on non-resonance, neither of which is established; the paper's own admission that the atom equivalence relation 'is yet to be explored' is a serious restriction on the announced scope.","major_comments":[{"comment":"The title and abstract promise a moduli theory of 'atoms' of Dubrovin connections of projective varieties, but this is not delivered. The reduction in §1 from quantum connections to objects of NSQC_c is heuristic: the stationary-phase decomposition (1) is asserted; the passage 'K_{−2} ... is the sum of a simple endomorphism and a nilpotent one. Up to simple endomorphisms, the study can therefore be reduced...' is not proved; and the non-resonance condition c_i ≠ c_j is never verified for actual quantum connections (e.g., Fano-type examples). The paper itself states that the atom equivalence relation 'is yet to be explored'. Hence Theorem 4.1 is, as written, a theorem about an abstract groupoid; the announced 'moduli of atoms of varieties' is not defined or proven. Either the reduction and non-resonance should be established (or precisely cited), or the paper should be reframed around NSQ","section":"§1"},{"comment":"The transformation law for d_{ji} appears to have the sign reversed relative to the gauge convention (5). For g = I + ℓ_{ji}E_{ji} (j<i) and diagonal part diag(μ_1/z,...,μ_r/z), the (j,i)-entry transforms as d_{ji} ↦ d_{ji} + (μ_j−μ_i)ℓ_{ji}/z + ∂_zℓ_{ji} = d_{ji} + L_{μ_i−μ_j}(ℓ_{ji}), not d_{ji} + L_{μ_j−μ_i}(ℓ_{ji}) as displayed. Lemma 4.4 and the conclusion A/Aut_0(E) ≅ n ⊕ n are unaffected if m(j,i) is redefined as max(μ_i−μ_j,0), so the theorem survives; nevertheless Prop 4.3 as stated is derived from the wrong sign and the inconsistency should be resolved.","section":"§4, Prop. 4.3"},{"comment":"Theorem 4.1 asserts an isomorphism of coarse moduli spaces as schemes, but the proof establishes only a bijection on isomorphism classes via the normal-form slice; the scheme structure, universal property, and behavior in families are not addressed. Indeed §5.1 makes the braid-group action conditional on 'supposing that the construction of Theorem 4.1 goes through in family', indicating the relative version is not in hand. To justify the theorem as stated, the authors should construct the moduli functor/stack and prove that the slice represents it, or state a point-set/analytic version with the appropriate caveat.","section":"§4, Thm 4.1; §5.1"}],"minor_comments":[{"comment":"The displayed formula ℓ_ii(z) = exp(∫_1^z (d_ii−μ_i)/ζ dζ) has the wrong sign if the goal is to make the i-th diagonal entry equal to μ_i dz/z: under (5) the entry transforms by d_ii + d log ℓ_ii, so the integrand should be (μ_i−d_ii)/ζ.","section":"§4"},{"comment":"Theorem 1.1 assumes only c_i ≠ c_j, whereas the proof of Theorem 4.1 (via Lemma 4.2) requires c_i ≠ 1. The standing convention c_i ∈ C^× \\ {1} makes this harmless, but the hypotheses of the two theorem statements should be aligned.","section":"Thm 1.1 vs Thm 4.1"},{"comment":"The r ≥ 3 part of Proposition 2.1 is only sketched ('We will just indicate the main steps... leave the reader to work out the details'), and the displayed expansion of ∇^r e_r is not derived. Since the proposition is used in the introduction's motivation, it should be proved in full, or the sketch expanded to a real proof, or the statement demoted to a remark.","section":"§2, Prop. 2.1"},{"comment":"In the proof of Lemma 3.4, formula (12) sums over j < i and omits the j = i term; the term is restored in (13). This is a typographical slip only, but should be corrected.","section":"§3, Lemma 3.4"},{"comment":"The claim that a stabilizer h of an object 'must be constant too by h∘∇ = ∇∘h' is asserted without proof; it does follow from parallel transport of h(1)=I_r (or from a Fuchs-type argument), but the justification should be stated. Also, the content of relation (3) ('quantum Leray–Hirsch') is not described, so the proposed atom equivalence cannot be checked.","section":"§1"}],"recommendation":"major_revision","confidential_remarks":"The core normal-form computation is honest, checkable, and, modulo the sign issue in Prop 4.3, sound; the main problem is scope: the paper promises a theory of atoms of projective varieties and delivers a theorem about an abstract groupoid. I recommend major revision: the authors should either prove (or cite precisely) the reduction of Dubrovin connections to (Cond1)–(Cond3) and verify non-resonance for examples, or reframe the title/abstract/introduction so that NSQC_c is the main object and atoms are explicitly future work. The sign discrepancy and the scheme-structure gap in Thm 4.1 must also be addressed. I would not recommend rejection, as the normal-form result is of independent interest and the issues appear repairable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague, the one thing to know: Theorem 4.1 is a genuine theorem about the groupoid NSQC_c, and the proof chain (Prop 3.1 → Prop 3.2 → Prop 4.3 → Thm 4.1) is in decent shape; I spot-checked the coker computation and the invariance argument and they are right. The second thing: the title is ahead of the content. The 'atoms of complex projective varieties' announced in the title and abstract is not proven. The atom equivalence relation is explicitly deferred ('is yet to be explored'), and the reduction from Dubrovin connections to NSQC_c is a heuristic in §1, not a theorem.\n\nWhat is new and good: a complete normal form for rank-r meromorphic connections with regular singularity, pole order at most 2, nilpotent leading coefficient, and non-resonant semisimple monodromy — entries α_ji z^{-2} + β_ji z^{m(j,i)-1}, with α_ji invariant. The moduli of the groupoid is n⊕n on each Z^r-indexed component. Lemma 4.4 is clean; the L_μ action z^n ↦ (n−μ)z^{n−1} is correct, and z^{-2} is genuinely outside the image. The specialization to Levelt's normal form for N=1 is the right sanity check. I did not find a fatal internal error.\n\nSoft spots, in proportion: the main one is the gap between the announcement and the proof. The stationary-phase decomposition (1) is asserted, not proved; the claim that K_{-2} is nilpotent 'up to simple endomorphisms' is not established; and the non-resonance condition c_i ≠ c_j is a hypothesis that is never verified for the motivating quantum connections. That is a real limitation on scope, not a flaw in Theorem 4.1 itself. If actual Dubrovin connections fail these conditions, the theorem describes an abstract groupoid rather than moduli of atoms of varieties.\n\nSmaller issues: the r ≥ 3 part of Proposition 2.1 is handed to the reader, and the induction in Lemma 3.4 is compressed. Both look repairable, and the rank-2 case is written out. There is also an apparent typo in the displayed formula for ℓ_ii in §4 — the integrand should be (μ_i/ζ − d_ii)dζ.\n\nBottom line: this is for people working on quantum connections, regular singularities, and moduli of connections. The core theorem deserves a serious referee. I would send it out, not desk-reject, and ask the referee to verify the sketchy proofs and to press the authors on the §1 reduction. A revised version that scales the title and intro to what is actually proved, or else constructs the atom relation, would be much stronger.","headline":"The groupoid moduli theorem is real and mostly checks out, but the paper's title over-promises: the atoms-of-varieties application is deferred, and the reduction to NSQC_c is heuristic.","tokens_in":14809,"tokens_out":3806,"would_cite":true,"duration_ms":38726,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14D20","14J33","53D45"],"pacs":[],"model":"deepseek-v4-flash","headline":"The moduli space of constituent pieces ('atoms') of Dubrovin connections is an affine space of dimension r(r−1) when monodromy eigenvalues are distinct and not 1.","keywords":["moduli of connections","Dubrovin connection","quantum cohomology","regular singularity","non-semisimple","normal form","mirror symmetry","birational invariants"],"falsifier":"Exhibit a smooth projective variety whose Dubrovin connection has either a non-nilpotent leading term K_{−2} (up to scalar) or resonant monodromy (some c_i=c_j or c_i=1); if its atom is not gauge-equivalent to the normal form α z^{−2}+β z^{m−1}, the theorem's geometric interpretation fails. Alternatively, for r=2 with c_1=1, c_2≠1, compute the coarse moduli space directly and show it is not an affine line.","tokens_in":13528,"feed_emoji":"📐","tokens_out":7281,"duration_ms":61547,"temperature":0.7,"pith_summary":"The paper introduces the groupoid NSQC_c of meromorphic connections on a rank-r vector bundle over a punctured disk, satisfying three conditions: regular singularity at 0, pole order at most 2, and monodromy that preserves a full flag with prescribed eigenvalues c_i. The main theorem states that if the c_i are distinct and none equals 1, each connected component of the coarse moduli space is isomorphic to the affine space n⊕n, of dimension r(r−1). The proof gives a normal form: any such connection is uniquely gauge-equivalent to one whose entries above the diagonal are α_{ji}z^{−2} + β_{ji}z^{m(j,i)−1}, so α, β are global coordinates. This is the first moduli statement for non-semisimple quantum connections, motivated by the goal of defining 'atoms' of Dubrovin connections as birational invariants of smooth projective varieties.","feed_headline":"Non-semisimple quantum connections get affine moduli","feed_subtitle":"With distinct monodromy eigenvalues, each moduli component is an affine space with explicit α,β coordinates.","key_machinery":"The central mechanism is the normal-form reduction of Propositions 3.2 and 4.3. The first step uses cyclic vectors and Fuchs' criterion to show that the connection matrix can be made polynomial, with the leading coefficient strictly upper triangular and the rest upper triangular. The second step uses the gauge group Aut_0(E) (upper-triangular holomorphic matrices with eigenvalue 1 at z=1) and the linear operators L_μ(P)=z^μ ∂_z(z^{−μ}P). The cokernel computation of L_μ, which is generated by z^{−2} and either z^{−1} (μ∉N) or z^{μ−1} (μ∈N), dictates the surviving coefficients, yielding the shape α z^{−2} + β z^{m−1}. The function m(j,i)=μ_j−μ_i if positive integer, else 0, determines which β","core_discovery":"The central claim is Theorem 4.1: for c_i distinct and not 1, the coarse moduli space of the groupoid NSQC_c has connected components isomorphic to n⊕n, the product of two copies of the nilpotent radical of a Borel. The proof proceeds by first showing (Proposition 3.2) that every object can be gauge-transformed to a polynomial connection with strictly upper triangular D_{−2} and upper triangular D_{−1}, ..., D_{N′}, using the existence of a monodromy-invariant flag and the regular singularity condition. Then (Proposition 4.3) a unique gauge transformation in Aut_0(E) brings each (j,i)-entry to α_{ji}z^{−2} + β_{ji}z^{m(j,i)−1}, where m(j,i) is μ_j − μ_i if this is a positive integer and 0 ot","pith_inferences":["If the normal form applies to actual quantum connections of Fano-type varieties, then the α, β parameters should be expressible in terms of Gromov–Witten invariants, making the atom a computable birational invariant.","The non-resonance condition c_i≠c_j is probably not essential: the cokernel of L_μ is still two-dimensional when μ is an integer or zero, so a generalized normal form with slightly modified β terms should hold in the resonant case, though the uniqueness argument would need adjustment.","The affine structure of the moduli space suggests that non-semisimple quantum cohomology, like the semisimple case, admits flat (Darboux-type) coordinates, which may simplify the study of isomonodromic deformations.","A concrete test would be to compute the atom for a specific non-semisimple Fano variety and check whether its quantum connection, after Fourier–Laplace transform, lies in NSQC_c and how many α, β parameters are needed."],"forward_implications":["Each object of NSQC_c is uniquely classified by the r(r−1) complex numbers (α_{ji}, β_{ji}), so the moduli space is a global affine chart with no topological obstructions.","The moduli space's connected components are indexed by a Z^r torsor, reflecting the choice of logarithm of the monodromy; within each component the space is a vector space.","The dimension r(r−1) is independent of the eigenvalues c_i, as long as they are distinct and not 1, so the atom of a variety has a fixed number of parameters for a given rank.","The normal form gives a concrete target for computing atoms of Dubrovin connections: for a given variety, the α, β parameters are determined by the connection's Laurent expansion.","The comparison with tame parahoric bundles suggests a Riemann–Hilbert-style description of these moduli spaces, a direction the paper explicitly flags for future work."],"fun_headline_variants":["Affine moduli for non-semisimple quantum connections","Distinct eigenvalue moduli: affine spaces with α,β","Moduli of Dubrovin atoms become affine spaces","Quantum connection moduli: components are n⊕n","Gauge normal form gives affine moduli components"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The proof assumes that every Dubrovin connection of a smooth projective variety, after the standard stationary-phase reduction, satisfies the three conditions of NSQC_c — regular singularity, pole order at most 2, and monodromy with distinct eigenvalues not equal to 1 — but this is only heuristically argued, not verified for the motivating examples.","fun_headline_variants_meta":{"raw":{"variants":["Affine moduli for non-semisimple quantum connections","Distinct eigenvalue moduli: affine spaces with α,β","Moduli of Dubrovin atoms become affine spaces","Quantum connection moduli: components are n⊕n","Gauge normal form gives affine moduli components"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000286,"raw_usage":{"total_tokens":1429,"prompt_tokens":565,"completion_tokens":864,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":309,"completion_tokens_details":{"reasoning_tokens":786}},"tokens_in":309,"tokens_out":864,"duration_ms":8408,"temperature":1.0,"reasoning_tokens":786,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T05:54:55.982325+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Exhibit a smooth projective variety whose Dubrovin connection has either a non-nilpotent leading term K_{−2} (up to scalar) or resonant monodromy (some c_i=c_j or c_i=1); if its atom is not gauge-equivalent to the normal form α z^{−2}+β z^{m−1}, the theorem's geometric interpretation fails. Alternatively, for r=2 with c_1=1, c_2≠1, compute the coarse moduli space directly and show it is not an affine line.","supporting_citations":[],"review_version":1}