{"id":"71c9bb3e-189f-4e04-bb5e-c60e8f572abc","arxiv_id":"2501.15607","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Conjectures II and III extend the Gromov-Witten/Pairs descendent correspondence to families of 3-folds, covering diagonal and general descendent insertions.","lead":"Rahul Pandharipande's lecture note lays out a conjectural framework equating two ways of counting curves in families of 3-dimensional spaces. If the new conjectures hold, they unify Gromov-Witten and stable pairs descendent theories for families, going beyond earlier proofs for toric 3-folds and the quintic.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The conjectures' well-posedness depends on an unproved rationality of stable pairs descendent series in q; without it, the substitution -q=e^{iu} in Conjectures I-III is not a defined operation, as the paper concedes in Section 3.3.","rationale":"The reader's weakest assumption identifies the same load-bearing point: the substitution -q=e^{iu} is only meaningful if the stable pairs descendent series is rational in q, and the paper itself states that the well-posedness of the correspondence depends on this conjecture. This is indeed where the central claim is least secure. The paper is explicitly a conjecture paper, so absence of proof for the main equality is not by itself a flaw; however, the rationality condition is not merely an unproved technical lemma but a prerequisite for the statements of Conjectures II and III to be well-formed. I do not see an internal inconsistency in the formulas of Section 5: the sign discussion is plausible, and the specialization of Conjecture III to Conjectures I and II appears coherent. The reliance on the matrix ~K from [23] is acceptable because it is a cited construction with stated properties. Therefore the appropriate verdict remains CONDITIONAL, with the condition already made explicit by the paper and by the reader.","tokens_in":12494,"tokens_out":10467,"duration_ms":102101,"concrete_test":"Pick the simplest non-toric family where the no-insertion correspondence is proven, the relative boundary family nu: X = C^2 x C -> M_{g,n} of Theorem 5, and compute the stable pairs descendent series for a diagonal insertion, e.g., tau_{0,0}(gamma * Delta_2) with gamma a point class on C^2, in low Euler characteristic. If the resulting q-series is not a rational function in q, the substitution -q=e^{iu} in Conjecture II is undefined; if it is rational and matches the Gromov-Witten side order by order in u, the rationality assumption gains a nontrivial check outside the Kunneth/toric cases.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central new content is the family-level conjectural equivalence of stable pairs and Gromov-Witten descendent series (Conjectures II and III, Sections 5.2-5.4). Both statements transform the stable-pairs variable q via -q = e^{iu}. But the stable-pairs series is defined only as an element of H*(Y)((q)) (Section 2.5); substituting q = -e^{iu} is not a formal operation on Laurent series. The paper explicitly flags this in Section 3.3: 'the stable pairs descendent series is conjectured to be a rational function in q... The well-posedness of the descendent correspondence depends upon the rationality conjecture.' Conjectures II and III repeat the same reliance (Sections 5.2, 5.4). No proof or independent evidence for rationality of diagonal or general descendent series is offered beyond the cases where Conjecture I is already known and a Kunneth decomposition is available. Thus the headline equality is not currently a well-formed assertion of equality of Laurent series; it is a conditional statement whose condition is both unproved and essential. This is the most load-bearing weakness because if rationality fails in any family with non-factorizable insertions, Conjectures II and III have no literal meaning, and Conjecture I is restricted to the established cases. The concern is not that the conjecture is false, but that it is not yet a fully defined mathematical statement in the family generality claimed.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper is an expository and programmatic article on the Gromov-Witten/stable pairs descendent correspondence in families of 3-folds. The author defines descendent series for stable maps and stable pairs on a family ν: X → Y, recalls the universal correspondence matrix ~K constructed with Pixton, and states Conjecture I (the standard descendent correspondence), Conjecture II (a symmetric form with diagonal descendents on both sides), and Conjecture III (the most general form with arbitrary insertions δ ∈ H*(X^ℓ)). The paper also surveys known cases: toric 3-folds, the quintic, log Calabi-Yau pairs, and primary insertions in Fano/Calabi-Yau families, and mentions the connection to the crepant resolution conjecture for Hilbert schemes of points of C^2. The central new content is the formulation of Conjectures II and III.","tokens_in":12747,"tokens_out":5809,"duration_ms":51680,"significance":"If Conjectures II and III hold, they would establish a complete equivalence between descendent Gromov-Witten and stable pairs theories for families of 3-folds at all genera and with arbitrary insertions, a major structural result in enumerative geometry. The paper is transparent about the conjectural status: it explicitly notes that the change of variables -q=e^{iu} is only well-defined after a rationality conjecture for the stable pairs series, and it limits the known family cases to situations with Kunneth decompositions and established Conjecture I. The formulation of the universal matrix ~K with its explicit properties is a useful organizing principle. The main value of the paper is as a precise conjectural framework and survey.","major_comments":[{"comment":"The statements of Conjectures I, II, and III are not well-formed as written because each asserts an equality between a stable pairs series, which is defined only as a Laurent series in q with coefficients in H*(Y), and a Gromov-Witten series in u, under the substitution -q=e^{iu}. As the paper itself notes in §3.3, this substitution is only a defined operation if the stable pairs series is a rational function in q, and the rationality of the general descendent series is an open conjecture. Consequently, the displayed equalities in Conjectures I, II, and III are conditional statements, not unconditional conjectures. The paper should restate each conjecture with an explicit hypothesis of the form 'Assume the rationality of the relevant stable pairs descendent series' and should state the conditional implication rather than presenting the equality as an unconditional assertion. This is a load-bearing point because if rationality fails, the claimed equality has no literal meaning.","section":"§3.3, §5.2, §5.4"},{"comment":"In Conjecture III, the sign conventions for odd cohomology classes are not specified. The text states that 'a nice exercise is to specialize Conjecture III to the cases of Conjectures I and II and to derive the sign rules there from Conjecture III,' but the statement of Conjecture III itself should be independent of such an exercise. Without an explicit sign rule or a precise reference to where the signs are defined, the right-hand side of the correspondence rule for general δ ∈ H*(X^ℓ) is not fully determined, and the conjecture is not completely well-posed. Please provide the sign convention for Conjecture III or rewrite the statement so that the specialization to Conjectures I and II is immediate.","section":"§5.4"}],"minor_comments":[{"comment":"The phrase 'cohomology 1 classes' should read 'cohomology classes'.","section":"§2.2"},{"comment":"The displayed formula for the fiber product X^r contains corrupted text ('bracehtipupleft /bracehtipdownright/bracehtipdownleft /bracehtipupright') that should be cleaned up before publication.","section":"§2.3"},{"comment":"There is a spacing typo in 'Chern cha racter'; it should be 'Chern character'.","section":"§2.4"},{"comment":"The definition of the meet D∧P of two set partitions is terse; a short concrete example illustrating the construction would improve readability.","section":"§5.2"}],"recommendation":"major_revision","confidential_remarks":"This is an invited lecture article and the heavy self-citation is natural given that the framework is the author's own. The main technical concern, namely the conditional nature of the conjectures on an unproved rationality statement, is explicitly acknowledged in the text but should be built into the formal statements of the conjectures. The sign issue in Conjecture III is a smaller but still substantive gap in well-posedness. Both are fixable within the scope of the manuscript."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The headline: this is a short lecture-note paper that does two things. It surveys the GW/P descendent correspondence as developed with Pixton, and it states two new conjectures, II and III, that extend the correspondence to diagonal and general descendent insertions over arbitrary smooth families of 3-folds. The new conjectures are the genuine content. They are clearly formulated, reduce to Conjecture I in the right special cases, and Section 5.4 is honest that they are only known in cases where Kunneth decompositions are available and Conjecture I is proven.\n\nThe paper does its job well. The survey is compact and useful. The notation is heavy but precise. The author is explicit about the main caveat: the stable pairs side is a Laurent series in q, and the change of variables -q = e^{iu} is not formal; it requires the rationality conjecture for the descendent series. That is flagged in Section 3.3 and repeated in 5.2 and 5.4. So the stress-test concern is real but it is not hidden. The conjectures are conditional on an unproved rationality statement, and the paper says so.\n\nThe soft spots, in proportion: first, the conditional nature is not a minor footnote. Without rationality, Conjectures II and III are not well-defined equalities of series. The paper's own statement that well-posedness depends on it is correct. A reader should not come away thinking the correspondence is a theorem in any generality beyond the known toric/quintic cases. Second, there is no new evidence for rationality beyond the cases where Conjecture I already holds; the general family case is genuinely open. Third, the dependence on the matrix ~K from [23] is heavy, but that is legitimate: ~K is a constructed object with proven properties, not a black box. The self-citation is heavy but appropriate for a research program.\n\nNone of this undercuts the value. The paper is a precise statement of what the full correspondence should be, and that is what a conjecture paper is for. I would send it to a serious referee, and I would expect the referee to check the sign conventions in Section 5.3 and the reduction of Conjecture III to I and II.\n\nFor a reader working on curve counting or stable pairs, this is worth having on the desk. I would cite it.","headline":"A clear, honest conjecture paper: new Conjectures II and III extend the GW/P descendent correspondence to general insertions in families, but their well-posedness depends on an unproved rationality conjecture that the paper itself acknowledges.","tokens_in":13311,"tokens_out":2345,"would_cite":true,"duration_ms":20874,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14N35","14D20","14J32"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper conjectures that for every smooth projective family of 3-folds, the descendent Gromov-Witten and stable pairs curve-counting theories are equivalent after the substitution -q = e^{iu}.","keywords":["Gromov-Witten theory","stable pairs","Donaldson-Thomas theory","descendent correspondence","families of 3-folds","moduli of sheaves","enumerative geometry","Calabi-Yau threefolds"],"falsifier":"A single concrete instance where the stable pairs descendent series is not rational in q, or where the two series in Conjecture III differ after the prescribed change of variables and factors, would falsify the correspondence.","tokens_in":12245,"feed_emoji":"📐","tokens_out":7888,"duration_ms":65732,"temperature":0.7,"pith_summary":"This paper presents a conjectural framework, in the context of families of 3-folds, for the Gromov-Witten/Pairs descendent correspondence: the claim is that two different ways of counting curves — via stable maps (Gromov-Witten theory) and via stable pairs (Donaldson-Thomas theory) — produce equivalent descendent invariants after the change of variables -q = $e^{{iu}}$, up to explicit factors. The paper states three nested conjectures: Conjecture I for standard descendent insertions, Conjecture II for diagonal descendent insertions, and Conjecture III for arbitrary cohomology classes on X^ℓ. If these conjectures hold, the entire descendent theories of the family match at all genera, making the two counting formalisms two lenses on the same underlying curve-counting data. The paper also surveys proven cases, including toric 3-folds, the quintic threefold, and primary insertions for Fano and Calabi-Yau fibers, and indicates how the correspondence connects to Virasoro constraints, the Crepant Resolution Conjecture, and moduli of curves.","feed_headline":"Conjectures unify two curve-counting theories for 3-fold families","feed_subtitle":"If all three hold, Gromov-Witten and stable pairs invariants match at every genus after -q = e^{iu}.","key_machinery":"The load-bearing object is the universal correspondence matrix $\\tilde{K}$, indexed by partitions $\\alpha$ and $\\hat\\alpha$, with entries in $\\mathbb{Q}[i, c_1, c_2, c_3]((u))$. It is constructed from the 1-legged capped descendent vertex for stable pairs and stable maps, and it has three defining properties: vanishing unless $|\\alpha| \\ge |\\hat\\alpha|$, homogeneity in the Chern classes $c_k$ with a specified degree, and a leading term that matches the identity on descendents up to a factor $(iu)^{\\ell(\\alpha)-|\\alpha|}$. The matrix furnishes a systematic rule that rewrites a product of standard descendent insertions as a sum over set partitions of diagonal descendent insertions; this rule is then used to state the correspondence after the change of variables $-q = e^{iu}$, where rationality of the stable pairs series is required for the substitution to be well-defined.","core_discovery":"The central claim of the paper is the 'GW/P families descendent correspondence': for a smooth projective family ν: X → Y of 3-folds, a fiber class β, and an insertion δ ∈ H*(X^ℓ), the stable pairs partition function $Z_P\\big(\\nu; q \\mid \\tau_{k_1,\\dots,k_\\ell}(\\delta)\\big)_\\beta$ and the disconnected Gromov-Witten partition function $Z'_{GW}\\big(\\nu; u \\mid \\tau_{k_1,\\dots,k_\\ell}(\\delta)\\big)_\\beta$ satisfy $$(-q)^{-d_\\$\\beta$/2}\\,Z_P = (-iu)^{d_\\$\\beta$}\\, Z'_{GW}$$ under the variable change $-q = e^{iu}$, where $d_\\beta = \\int_\\beta c_1(T_\\nu)$. Conjecture I covers standard descendent insertions (products of $\\tau_{\\alpha_i-1}(\\gamma_i)$), Conjecture II extends to diagonal descendent insertions built from the small diagonal and the correspondence matrix $\\tilde{K}$, and Conjecture III handles arbitrary classes $\\delta \\in H^*(X^\\ell)$ via a more intricate set-partition formula. The paper's evidence and the formulation of all three conjectures rest on the universal correspondence matrix $\\tilde{K}$, constructed from the capped descendent vertex, which converts products of standard descendents into sums of diagonal descendents with coefficients in $\\mathbb{Q}[i, c_1, c_2, c_3]((u))$.","pith_inferences":["The rationality conjecture for stable pairs descendent series, if true, would make the change of variables $-q = e^{iu}$ an actual analytic continuation, hinting at a deeper modular or q-analogue structure relating the two theories.","The universal matrix $\\tilde{K}$ might be interpreted as a change-of-basis between two cohomological field theories; a motivic or categorical refinement could lift the numerical equality to an equality of virtual motive-valued invariants.","Because the conjectures are stated in families, they suggest the correspondence is a sheaf-theoretic statement over the base $Y$, not only a pointwise identity, with potential consequences for enumerative geometry over moduli spaces of curves.","A concrete open test would be to establish Conjecture II for the universal family over $M_{g,n}$ with one descendent insertion, which would produce new constraints on the tautological cohomology of the moduli space of curves."],"forward_implications":["If Conjecture III holds, the descendent Gromov-Witten and stable pairs theories of any smooth projective family of 3-folds are equivalent at all genera with arbitrary insertions, yielding a single unified curve-counting invariant.","The proven toric case already implies Virasoro constraints for moduli spaces of sheaves on toric 3-folds, and the full conjecture would extend these constraints to broader classes of 3-folds.","Conjecture II specializes to Conjecture I, so proving the symmetric diagonal form would automatically settle the standard descendent correspondence for every family.","The correspondence for the universal curve family $C^2 \\times C \\to M_{g,n}$ is equivalent to the Crepant Resolution Conjecture for Hilbert schemes of points of the plane; the stronger conjectures would refine this to descendent insertions.","The recent proof of the primary-insertion case for Fano and Calabi-Yau families is a step toward the full conjecture, which would subsume that result."],"supporting_citations":[{"why":"Constructs the correspondence matrix ~K and proves the descendent correspondence for toric 3-folds, supplying the core machinery on which Conjectures I–III are built.","marker":"[23]"},{"why":"Proves the correspondence for the quintic threefold via a degeneration strategy, the main evidence that the conjecture holds beyond toric geometries.","marker":"[24]"},{"why":"Provides the foundational moduli space of stable pairs and the descendent insertions used on the pairs side of the correspondence.","marker":"[25]"},{"why":"Formulates the original Gromov-Witten/Donaldson-Thomas correspondence for ideal sheaves, the ancestor of the descendent correspondence studied here.","marker":"[9, 10]"},{"why":"Proves the primary-insertion correspondence for toric 3-folds, the base case that the family version extends.","marker":"[11]"},{"why":"Proves the primary-insertion correspondence for families with Fano or Calabi-Yau fibers using new transversality arguments, a recent step the conjectures would generalize.","marker":"[29]"}],"fun_headline_variants":["Three conjectures unify curve and sheaf counts on 3-fold families","Family GW/DT correspondence conjectured with three cases","Curve-sheaf equivalence conjectured for families of 3-folds"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The stable pairs descendent partition function is assumed to be a rational function in q, so that substituting -q = $e^{{iu}}$ is well-defined; if rationality fails for some family or insertion, Conjectures I–III are not well-formed.","fun_headline_variants_meta":{"raw":{"variants":["Three conjectures unify curve and sheaf counts on 3-fold families","Family GW/DT correspondence conjectured with three cases","Curve-sheaf equivalence conjectured for families of 3-folds"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000831,"raw_usage":{"total_tokens":3654,"prompt_tokens":999,"completion_tokens":2655,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":615,"completion_tokens_details":{"reasoning_tokens":2595}},"tokens_in":615,"tokens_out":2655,"duration_ms":17743,"temperature":1.0,"reasoning_tokens":2595,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T14:07:00.708826+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A single concrete instance where the stable pairs descendent series is not rational in q, or where the two series in Conjecture III differ after the prescribed change of variables and factors, would falsify the correspondence.","supporting_citations":[{"cited_title":"Pandharipande and A","cited_arxiv_id":null,"evidence_quote":"Constructs the correspondence matrix ~K and proves the descendent correspondence for toric 3-folds, supplying the core machinery on which Conjectures I–III are built."},{"cited_title":"Pandharipande and A","cited_arxiv_id":null,"evidence_quote":"Proves the correspondence for the quintic threefold via a degeneration strategy, the main evidence that the conjecture holds beyond toric geometries."},{"cited_title":"Pandharipande and R","cited_arxiv_id":null,"evidence_quote":"Provides the foundational moduli space of stable pairs and the descendent insertions used on the pairs side of the correspondence."},{"cited_title":"Maulik, A","cited_arxiv_id":null,"evidence_quote":"Proves the primary-insertion correspondence for toric 3-folds, the base case that the family version extends."}],"review_version":1}