{"id":"c60d4c79-004f-4151-a6cc-8773d385ee03","arxiv_id":"2608.04795","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The author introduces Joyce invariants for Quot schemes on curves, proves a wall-crossing formula and Virasoro constraints, and gives a recursive method to eliminate f-classes from virtual intersections.","lead":"This paper defines a Joyce-type enumerative invariant for Quot schemes of vector bundles on curves and proves Virasoro constraints for their virtual intersection theory. It then uses these constraints to reduce integrals of f-classes to a- and b-class integrals.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 5.9's wall-crossing pushforward assumes, without verification, that Ξ commutes with the Lie bracket for the e=1 point class; the cited K-theoretic restriction to M×M does not cover the brackets used in (22).","rationale":"The central claim of the paper is Theorem 5.9: all descendents on Quot_{r,d}(E) satisfy Virasoro constraints, which would give universal linear relations among tautological intersection numbers and eliminate f-classes. The proof must show that the class Quot_{r,d}(E) is primary in the pair vertex algebra. That conclusion rests on two things: the wall-crossing formula (10) surviving pushforward by Ξ, and the external theorem [BLM24, Thm 5.12] that the sheaf invariants M_α are primary states. The reader's weakest assumption named both, but the more concrete internal weakness is the bracket compatibility. The paper's one-sentence justification for compatibility checks only \\check{M}×M, i.e., one factor in the sheaf subalgebra V_* (e=0), whereas the brackets actually used in (22) pair M_α with the e=1 point class e_(0,0,E). The state-field correspondence in (16) contains terms involving V and the E^∨⊠V contribution from the \\check{Ext} complex, so K-theoretic equality on \\check{M}×M does not automatically extend to the needed component. This is a finite, checkable computation from the explicit formulas, so the concern is concrete rather than a vague appeal to rigor. If the compatibility fails, the wall-crossing formula cannot be pushed to qV^pa and the primary-state argument collapses. If it passes, the proof is plausible modulo the stated BLM24 inputs. The rank N−1 computation in Section 4 and the N^{g−s} formula in Theorem 6.7 provide some independent consistency evidence for the wall-crossing machinery in a special case, but they do not exercise the e=1 bracket compatibility needed for general r. In particular, the sign cancellation between rP = −[e,M] and the sign in Res Y(e,−z) appears to work out, but the Virasoro proof still lacks the missing compatibility check.","tokens_in":35513,"tokens_out":20349,"duration_ms":227035,"concrete_test":"Restrict \\check{Ext} and (Ξ×Ξ)^*Ext^pa to the component \\check{M}_{(0,0),1}×M_α and compute the residues defining [e_(0,0,E), M_α] in (7) using (16) and the analogous pair-vertex-algebra formula in [BLM24, §4.3]; compare the result with [Ξ_*e, Ξ_*M_α] in qV^pa. A nonzero difference, or a sign or coefficient mismatch in the surviving class e_(α,1), would invalidate formula (22) and with it the proof of Theorem 5.9.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The proof of Theorem 5.9 pushes the wall-crossing formula (10) along Ξ^pl to qV^pa, then lifts it to V^pa and applies [BLM24, Thm 5.12] that the sheaf invariants M_α are primary. The pushforward step requires that the iterated Lie brackets in (10), which involve e_(0,0,E) and the M_α_i, coincide with the brackets in qV^pa. The paper justifies this by saying that the restrictions of the complexes \\check{Ext} and (Ξ×Ξ)^*Ext^pa on \\check{M}×M⊂\\check{M}×\\check{M} coincide as K-theory classes, and concludes that Ξ_*(\\check{Y}(u,z)v)=Y^pa(Ξ_*u,z)(Ξ_*v) for u∈\\check{V}_*, v∈V_*. But the class e_(0,0,E) has e=1 and lies in \\check{M}_{(0,0),1}, which is not in the subalgebra V_* (those classes have e=0). Hence the stated hypothesis v∈V_* fails for every bracket involving e_(0,0,E). The required compatibility for u∈\\check{V}_{(0,0),1}, v∈V_{*,α} (or the reverse order) is not established. If the state-field correspondences differ on this component, formula (22) does not hold in qV^pa, and the subsequent lift via Lemma 5.8(2) is unjustified. This is an internal gap in the proof of the central theorem, not merely a deferral to a cited result.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper applies Joyce's enumerative wall-crossing framework to Quot schemes of a fixed vector bundle on a smooth projective curve. It defines a Joyce invariant for Quot_{r,d}(E), proves a wall-crossing formula expressing it in terms of the point class e_{(0,0),E} and Joyce invariants of sheaf moduli, computes the invariant explicitly in the rank-(N-1) case, and uses it to recover intersection numbers of a- and b-classes. It then formulates Virasoro constraints for Quot schemes following BLM24, proves them in Theorem 5.9, and derives from them a recursive elimination of f-classes in tautological intersection numbers. The paper also contains explicit low-dimensional examples and a comparison with the known count N^{g-s} of maximal subbundles.","tokens_in":35890,"tokens_out":12237,"duration_ms":141912,"significance":"If the main theorem is correct, the paper provides a substantial extension of the sheaf-theoretic Virasoro constraints of BLM24 to Quot schemes, giving universal linear relations among descendent integrals and a method to eliminate f-classes. The explicit wall-crossing formula for Joyce's invariant of Quot schemes and the closed-form computation for rank-one kernels are useful contributions. The paper deserves credit for checking its machinery against an independent baseline: Theorem 6.7 reproduces the known number N^{g-s} of maximal subbundles, and Section 6 derives concrete f-class relations. The main caveat is that the proof of Theorem 5.9 has a gap in the compatibility of the wall-crossing bracket with the pushforward to the pair vertex algebra, and several foundational statements are proved only by reference to Joyce or BLM24.","major_comments":[{"comment":"The proof asserts that the point class e_{(0,0),E} can be pushed forward through Ξ and that the iterated Lie brackets in formula (22) are compatible with the pair vertex algebra bracket. The justification is that the restrictions of \\check{Ext} and (Ξ×Ξ)^*Ext^pa agree on \\check{M}×M, and hence state-field correspondences coincide for u∈\\check{V}_* and v∈V_*. However, e_{(0,0),E} has e=1 and lies in \\check{M}_{(0,0),1}, which is not contained in V_* (whose classes have e=0). Thus every bracket involving e_{(0,0),E}, in either order, violates the stated hypothesis v∈V_*. Since formula (10) and its rewrite (22) use exactly those brackets, the equality in qV^pa and the subsequent lift via Lemma 5.8(2) are not justified as written. This is an internal gap in the proof of the central theorem, not merely a deferral to a cited result.","section":"Theorem 5.9, proof"},{"comment":"Theorem 3.2 is the foundation of the entire wall-crossing construction, but its proof consists of saying that Joyce's Assumptions 4.4 and 5.1-5.3 \"follow similarly\" to [Joy21, Sections 8.2.1-8.2.4] after replacing the line bundle with a vector bundle. This is load-bearing because Theorem 3.5 and formula (10) depend on it. Please provide a detailed verification of the finiteness, permissible-class, and stability assumptions for E^∨-pairs for arbitrary vector bundles, or state precisely which results in [Joy21] already cover this case.","section":"Theorem 3.2"},{"comment":"The displayed recursive formula does not appear to follow from (24) and (25). Substituting (25) into S_{1,2,l+1} from (24) with m=l gives expressions involving μ_{l+1} and f_i for i≤l+1, but the right-hand side of (27) contains only ∑_{i=1}^l ∂μ_l/∂a_i f_i and a second derivative sum over i,k≤l+1 of μ_l, which is not defined. As written, the formula cannot be used recursively to eliminate f_{l+1}. Please correct the indices and clarify the convention for μ_0 and μ_l, or explain if a different indexing is intended.","section":"Theorem 6.3, Eq. (27)"}],"minor_comments":[{"comment":"The graded vertex algebra structure is invoked from [Joy21] before the grading shift has been fully explained for the reader; a short reminder of the parity convention would improve readability.","section":"Section 3, Theorem 3.1"},{"comment":"The pairing formula is difficult to parse because the ranges of the products and the meaning of the total order are compressed; please spell out the indexing more explicitly.","section":"Equation (14)"},{"comment":"The stability condition μδ is defined only for rank F > 0; for e=0 pairs and for F=0 the reader must infer the limiting convention. Please state the definition uniformly.","section":"Section 2, stability definition"},{"comment":"The class S_{1,0,0} is used implicitly in the Virasoro relation but is never defined; please state that S_{1,0,0}=rank K^∨ under the geometric realization.","section":"Example 6.4"},{"comment":"The residue computation ends with \"This can be easily calculated to be N^{g-s}\"; since this is the only place the numerical factor N^{g-s} is obtained, please include the intermediate steps.","section":"Theorem 6.7"},{"comment":"Reference [Joy19] is cited by a URL; if a stable published or arXiv version exists, please cite it in addition.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The proof of Theorem 5.9 has a genuine gap concerning the e=1 point class, and the formula in Theorem 6.3 has an index inconsistency. Both are likely fixable, but as written the main theorem is not fully proved. The paper's reliance on [BLM24] and [Bu23] is heavy but acceptable in principle; the issue is the missing verification at the specific bracket compatibility step."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: the paper has a genuinely new target for the Virasoro program — Quot schemes on curves — and a useful new computation in rank N-1, but the proof of the central theorem has an internal gap that needs to be closed before the headline result can be trusted.\n\nWhat is actually new: the wall-crossing formula for Joyce's invariant on Quot schemes, the explicit invariant for Quot_{N-1,d}(E), the f-class elimination recursion, and the observation that the Virasoro formalism applies to this setting. Section 4's computation is concrete and is checked against an independent known count (N^{g-s} in Theorem 6.7), so that part is credible. The exposition is transparent; the author says plainly which results are taken from Joyce and BLM24, and the citations are appropriate.\n\nThe soft spot is in Theorem 5.9, and the stress-test note is on target. The pushforward compatibility between the E-pair vertex algebra and the pair vertex algebra is only established for a state-field product Y(u,z)v when v lies in the sheaf subalgebra V_* (e=0). The wall-crossing bracket starts with e_(0,0,E), which has e=1, and after the first bracket the intermediate state also has e=1. So the hypothesis v∈V_* fails for every bracket beyond the first, and the paper does not prove compatibility for the e=1 components. That is an internal gap in the argument, not just a deferral to a cited theorem. It may well be repairable by proving the e=1 compatibility directly, but as written formula (22) is not justified.\n\nA smaller complaint: the abstract says f-intersections are computed; the paper actually gives a recursive elimination algorithm, and it says so in the body. That is fine, but the abstract should match.\n\nAudience: people who work on tautological intersection theory for sheaf moduli and on vertex-algebra Virasoro constraints. The paper deserves a serious referee rather than a desk rejection: the intended application is substantial and the explicit computations are likely to be useful even if Theorem 5.9 needs work. I would not cite the Virasoro constraints as a theorem until the e=1 compatibility is sorted out.","headline":"A genuinely new extension of Virasoro constraints to Quot schemes, with a strong rank N-1 computation, but the proof of the main theorem has an internal e=1 compatibility gap that needs fixing.","tokens_in":36388,"tokens_out":4144,"would_cite":false,"duration_ms":47646,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14N35","14D20","14C17","17B69","14H60"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper claims that Quot schemes on curves satisfy Virasoro constraints, giving universal linear relations among all their tautological intersection numbers and reducing every f-class intersection to a- and b-class intersections.","keywords":["Quot schemes","Joyce invariant","wall-crossing","Virasoro constraints","vertex algebra","virtual fundamental class","tautological intersection numbers","vector bundles on curves"],"falsifier":"Specialize Theorem 6.7 to E = O_C^N and a degree d with dim Q ≥ 0: the theorem predicts ∫_{[Quot_{N-1,d}(O^N)]^{vir}} $a_1^{{dim Q}}$ = N^g. A direct virtual-localization computation of this single integral, using the torus action on O^N, that produces a value different from N^g would disprove the claimed Virasoro constraints and the associated f-class elimination.","tokens_in":35297,"feed_emoji":"📐","tokens_out":6959,"duration_ms":81233,"temperature":0.7,"pith_summary":"This paper claims that the Quot scheme Quot_{r,d}(E) of quotients of a vector bundle E on a smooth projective curve satisfies Virasoro constraints, i.e., universal linear relations among all tautological intersection numbers. To get there, the paper introduces Joyce's enumerative invariant for the Quot scheme, shows that it equals the pushforward of the virtual fundamental class, and proves a wall-crossing formula expressing this invariant through moduli of sheaves. The invariant is computed explicitly in the rank-one-kernel case, yielding intersection numbers of a- and b-classes. If the central theorem is correct, every f-class intersection on any such Quot scheme can be eliminated in favor of a- and b-class intersections by an explicit recursive formula.","feed_headline":"Virasoro constraints proven for Quot schemes on curves","feed_subtitle":"Joyce's wall-crossing invariant reduces every f-class intersection to a,b-classes.","key_machinery":"The load-bearing object is Joyce's invariant [Quot_{r,d}(E)]^{inv}, a class in the shifted homology of the stack of E^∨-pairs; it coincides with the pushforward f_*[Quot_{r,d}(E)]^{vir} and satisfies a wall-crossing formula as an iterated Lie bracket of a point class and sheaf invariants M_α. The Virasoro proof uses the pair vertex algebra $V^{{pa}}$ on the homology of pairs of complexes, whose conformal element produces operators $L_m^{{pa}}$ dual to the descendent Virasoro operators; the Quot class is shown to be a primary state, meaning all positive Virasoro operators kill it. The rank-(N-1) computation uses an explicit description of the homology algebra and the residue formula res_{z=0} $z^{{-(ν+N d_)}}$ ρ(z) σ(N/z - s_{1,2,2}).","core_discovery":"The central result is Theorem 5.9: for any m ≥ 0 and any descendent D, the integral of ξ_{K^∨}(L_m(D)) over the virtual fundamental class [Quot_{r,d}(E)]^{vir} vanishes. This means the Quot scheme satisfies the Virasoro constraints, so all descendent integrals obey universal linear relations. The paper further proves that any top-degree polynomial in the tautological classes a, b, and f can be replaced by a polynomial in a and b alone, with a recursive formula for the replacement. For the case where the kernel has rank one, it computes the invariant explicitly and obtains the intersection formula ∫ ∏_{i=1}^s b_{j_i}^1 b_{j_i+g}^1 $a_1^{{dim Q - s}}$ = $N^{{g-s}}$, where N = rank(E).","pith_inferences":["The recursive elimination of f-classes suggests that, once any computation of a- and b-intersections is available (for example by virtual localization), the full set of tautological intersection numbers on Quot schemes becomes accessible; the paper itself does not pursue this combination.","The explicit residue computation for rank-one kernels indicates that analogous closed-form intersection numbers for higher-rank kernels would follow if a similar invariant computation could be carried out, which the author leaves as future work.","The wall-crossing expression may be invertible, potentially allowing the invariant for arbitrary r to be built from lower-rank Quot invariants and sheaf invariants, a route that would bypass Virasoro constraints for explicit numerical evaluations.","The N^g count for maximal rank-one subbundles is stable under the choices in the wall-crossing formula; a natural testable extension is whether this count remains unchanged under deformations of E for higher-rank maximal subbundles."],"forward_implications":["Every descendent integral on Quot_{r,d}(E) obeys the Virasoro relations, so tautological intersection numbers satisfy universal linear constraints rather than being independent.","Any top-degree polynomial in the a, b, and f classes can be rewritten as a polynomial in a and b alone; the paper gives a recursive algorithm for the rewriting.","For rank-one kernels, all a/b intersections are explicit: the product formula ∫ ∏_{i=1}^s b_{j_i}^1 b_{j_i+g}^1 a_1^{dim Q - s} = N^{g-s} holds, and in dimension zero the virtual count of maximal rank-one subbundles is N^g.","The same wall-crossing framework proves Virasoro constraints for the related moduli space P(E,α) of stable pairs with sections.","With the invariant known for rank-one kernels and the Virasoro constraints in hand, complete tautological intersection numbers on those Quot schemes can be evaluated without computing f-class integrals directly."],"supporting_citations":[{"why":"Supplies the recipe for enumerative invariants, the vertex algebra structure on shifted homology, and the wall-crossing theorem used to define the Quot invariant.","marker":"[Joy21]"},{"why":"Proves sheaf-theoretic Virasoro constraints via the pair vertex algebra and supplies the primary-state theorem for M_α, the conformal element, and the Ad lift used in Theorem 5.9.","marker":"[BLM24]"},{"why":"Constructs the virtual fundamental class for Quot schemes and the perfect obstruction theory that Proposition 2.3 compares with the pair moduli space.","marker":"[MO07c]"},{"why":"Gives the explicit generators and algebra structure for homology of moduli stacks of complexes, used in the computation of the rank-one invariant.","marker":"[Gro20]"},{"why":"Computes Joyce's invariant in the line-bundle case, providing the computational template and explicit operator identities adapted in Section 4.","marker":"[Bu23]"},{"why":"Builds the derived moduli stacks of objects that form the ambient space for E^∨-pairs and for the pair vertex algebra.","marker":"[TV07]"},{"why":"Shows that a vertex algebra induces a Lie bracket on the quotient by the translation operator, which is how wall-crossing sums are interpreted as Lie brackets.","marker":"[Bor86]"}],"fun_headline_variants":["Virasoro constraints proven for curve Quot schemes","Joyce invariant solves Quot scheme intersections","Quot schemes satisfy Virasoro constraints on curves","All f-class intersections reduced to a,b classes by Joyce invariant","Explicit Quot scheme intersection numbers via Joyce's invariant"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof assumes that the sheaf-counting invariants appearing in the wall-crossing sum are primary states killed by all Virasoro operators, and that pushing the wall-crossing formula into the pair vertex algebra respects the Lie bracket; if either fails, the Virasoro vanishing for Quot schemes does not follow.","fun_headline_variants_meta":{"raw":{"variants":["Virasoro constraints proven for curve Quot schemes","Joyce invariant solves Quot scheme intersections","Quot schemes satisfy Virasoro constraints on curves","All f-class intersections reduced to a,b classes by Joyce invariant","Explicit Quot scheme intersection numbers via Joyce's invariant"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000241,"raw_usage":{"total_tokens":1508,"prompt_tokens":915,"completion_tokens":593,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":531,"completion_tokens_details":{"reasoning_tokens":516}},"tokens_in":531,"tokens_out":593,"duration_ms":6086,"temperature":1.0,"reasoning_tokens":516,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T16:23:23.752667+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Specialize Theorem 6.7 to E = O_C^N and a degree d with dim Q ≥ 0: the theorem predicts ∫_{[Quot_{N-1,d}(O^N)]^{vir}} $a_1^{{dim Q}}$ = N^g. A direct virtual-localization computation of this single integral, using the torus action on O^N, that produces a value different from N^g would disprove the claimed Virasoro constraints and the associated f-class elimination.","supporting_citations":[],"review_version":1}