Moduli of curves and moduli of sheaves

Rahul Pandharipande

REVIEW 2 major objections 4 minor 32 references

Moduli of curves and moduli of sheaves

T0 review · 2 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read 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}.

desk verdict 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. read the letter →

arxiv 2501.15607 v1 pith:I26STHSP submitted 2025-01-26 math.AG

Rahul Pandharipande This is my paper · ORCID

classification math.AG MSC 14N3514D2014J32

keywords Gromov-WittentheorystablepairsDonaldson-Thomasdescendentcorrespondencefamiliesof3-foldsmodulisheavesenumerativegeometryCalabi-Yauthreefolds

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

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.

What carries the argument

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.

What would settle it

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.

Watch

Extended reading notes

Core claim

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 $\mathbb{Q}[i, c_1, c_2, c_3]((u))$.

Load-bearing premise

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.

Editorial extensions

If this is right

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.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

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.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the 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.

The 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.

The 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.

None 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.

For a reader working on curve counting or stable pairs, this is worth having on the desk. I would cite it.

Referee Report

2 major / 4 minor

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.

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 (2)

[§3.3, §5.2, §5.4] 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.
[§5.4] 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.

minor comments (4)

[§2.2] The phrase 'cohomology 1 classes' should read 'cohomology classes'.
[§2.3] The displayed formula for the fiber product X^r contains corrupted text ('bracehtipupleft /bracehtipdownright/bracehtipdownleft /bracehtipupright') that should be cleaned up before publication.
[§2.4] There is a spacing typo in 'Chern cha racter'; it should be 'Chern character'.
[§5.2] The definition of the meet D∧P of two set partitions is terse; a short concrete example illustrating the construction would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the conjectures are stated conditionally and build on proven prior results, not on the target equality.

full rationale

The paper is a survey and conjecture paper. The central new statements (Conjectures II and III) are not derived from their inputs: the correspondence matrix ~K is constructed in prior work [23] from the capped descendent vertex, and the GW/P correspondence framework is carried over from [9,10,23,24]. Conjecture I is a conjecture, not a claim proven from ~K; Conjectures II and III are stated as generalizations via the correspondence rule, and their reduction to Conjecture I in special cases is a mathematical specialization, not a circular definition. The paper explicitly flags the load-bearing rationality assumption 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.' This is a genuine open condition, not a hidden use of the conclusion. Self-citations are frequent, but they cite prior proofs (toric case, quintic), which are independent evidence. No equation in the paper reduces by construction to its inputs, and no fitted parameter is renamed as a prediction. Hence no circularity.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

No free numerical parameters are fitted in this paper; the matrix ~K is a fixed object from [23]. Three axioms are load-bearing: the rationality conjecture, the correctness of ~K, and standard virtual class foundations. No invented entities are introduced.

assumptions (3)

domain assumption The stable pairs descendent series is a rational function in q, making the substitution -q = e^{iu} well-defined.
Section 3.3 states that the well-posedness of the descendent correspondence depends on the rationality conjecture.
domain assumption The universal correspondence matrix ~K exists and satisfies the vanishing, homogeneity, and leading-term properties (i) to (iii) from [23].
Section 3.2 builds the entire correspondence formula (3.3) from ~K, and no independent proof is given in this paper.
standard math Standard virtual class foundations for Gromov-Witten and stable pairs theories in families are correct.
Sections 2.2 and 2.4 cite [1, 8, 25] for deformation theory and virtual classes.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Moduli of curves and moduli of sheaves." pith.science (2026). https://pith.science/paper/I26STHSP

@misc{pith2026250115607,
  author       = {Pith},
  title        = {Pith review of: Moduli of curves and moduli of sheaves},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/I26STHSP}},
  note         = {Machine review of arXiv:2501.15607}
}

read the original abstract

Relationships between moduli spaces of curves and sheaves on 3-folds are presented starting with the Gromov-Witten/Donaldson-Thomas correspondence proposed more than 20 years ago with D. Maulik, N. Nekrasov, and A. Okounkov. The descendent and relative correspondences as developed with A. Pixton in the context of stable pairs led to the proof of the correspondence for the Calabi-Yau quintic 3-fold. More recently, the study of correspondences in families has played an important role in connection with other basic moduli problems in algebraic geometry. The full conjectural framework is presented here in the context of families of 3-folds. This article accompanies my lecture at the ICBS in July 2024.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

32 extracted references · 30 canonical work pages

[23]

Pandharipande and A

R. Pandharipande and A. Pixton, Gromov-Witten/Pairs descendent correspondence for toric 3-folds , Geom. Topol. 18 (2014), 2747–2821

work page 2014
[1]

Behrend and B

K. Behrend and B. Fantechi, The intrinsic normal cone, Invent. Math. 128 (1997), 45–88. Moduli of curves and moduli of sheaves 17

work page 1997
[2]

Bojko, W

A. Bojko, W. Lim, and M. Moreira, Virasoro constraints for moduli of sheaves and vertex algebras , Invent. Math. 236 (2024), 387–476

work page 2024
[3]

Bryan and T

J. Bryan and T. Graber, The crepant resolution conjecture , In: Alge- braic Geometry–Seattle 2005, Part 1, 23–42, Proc. Sympos. Pur e Math. 80, Amer. Math. Soc., Providence, RI, 2009

work page 2005
[4]

Bryan and R

J. Bryan and R. Pandharipande, Local Gromov-Witten theory of curves, JAMS 21 (2008), 101–136

work page 2008
[5]

Douady and J.-L

A. Douady and J.-L. Verdier, S´ eminaire de G´ eom´ etrie Analytique de l’Ecole Normale Sup´ erieure 1974/1975, Ast´ erisque36-37 (1976), Ex- pos´ es VI-IX

work page 1976
[6]

Givental, Gromov-Witten invariants and quantization of quadratic Hamiltonians, Moscow Math

A. Givental, Gromov-Witten invariants and quantization of quadratic Hamiltonians, Moscow Math. J. 1 (2001), 487–518

work page 2001
[7]

Le Potier, Faisceaux semi-stable et syst` emes coh´ erents, in Vector bun- dles in algebraic geometry (Durham, 1993) , LMS Lecture Note Ser

J. Le Potier, Faisceaux semi-stable et syst` emes coh´ erents, in Vector bun- dles in algebraic geometry (Durham, 1993) , LMS Lecture Note Ser. 208, 179–239, Cambridge Univ. Press: Cambridge, 1995

work page 1993

Show all 32 references

[8]

Li and G

J. Li and G. Tian, Virtual moduli cycles and Gromov-Witten invariants of algebraic varieties, JAMS 11, 119–174, 1998

1998
[9]

Maulik, N

D. Maulik, N. Nekrasov, A. Okounkov, and R. Pandharipande, Gromov- Witten theory and Donaldson-Thomas theory. I , Compos. Math. 142 (2006), 1263–1285

2006
[10]

Maulik, N

D. Maulik, N. Nekrasov, A. Okounkov, and R. Pandharipande, Gromov- Witten theory and Donaldson-Thomas theory. II , Compos. Math. 142 (2006), 1286–1304

2006
[11]

Maulik, A

D. Maulik, A. Oblomkov, A. Okounkov, and R. Pandharipande, The Gromov-Witten/Donaldson-Thomas correspondence for tori c 3-folds , Invent. Math. 186 (2011), 435–479

2011
[12]

Maulik and D

D. Maulik and D. Ranganathan, Logarithmic enumerative geometry for curves and sheaves , arXiv:2311.14150 (2023)

2023 arXiv
[13]

M. Moreira, Virasoro conjecture for the stable pairs descendent theory of simply connected 3-folds (with applications to the Hilbe rt scheme of points of a surface) , Journal of the LMS 106 (2020), 154–191

2020
[14]

Moreira, A

M. Moreira, A. Oblomkov, A. Okounkov, and R. Pandharipande, Vira- soro constraints for stable pairs on toric 3-folds , Forum of Mathematics Pi 10 (2022)

2022
[15]

Oberdieck and A

G. Oberdieck and A. Pixton, Holomorphic anomaly equations and the Igusa cusp form conjecture , Invent. Math. 213 (2018), 507–587

2018
[16]

Oblomkov, A

A. Oblomkov, A. Okounkov, and R. Pandharipande, GW/PT descen- dent correspondence via vertex operators , Comm. Math. Phys. 374 (2020), 1321–1359

2020
[17]

Okounkov and R

A. Okounkov and R. Pandharipande, Quantum cohomology of the Hilbert scheme of points of the plane , Invent. Math. 179 (2010), 523– 557

2010
[18]

Okounkov and R

A. Okounkov and R. Pandharipande, The local Donaldson-Thomas the- ory of curves , Geom. Topol. 14 (2010), 1503–1567

2010
[19]

Pandharipande, Descendents for stable pairs on 3-folds , Modern Ge- ometry: A celebration of the work of Simon Donaldson, Proc

R. Pandharipande, Descendents for stable pairs on 3-folds , Modern Ge- ometry: A celebration of the work of Simon Donaldson, Proc. Sympo s. 18 Rahul Pandharipande Pure Math. 99 (2018), 251–288

2018
[20]

Pandharipande and A

R. Pandharipande and A. Pixton, Descendents on local curves: Ratio- nality, Comp. Math. 149 (2013), 81–124

2013
[21]

Pandharipande and A

R. Pandharipande and A. Pixton, Descendents on local curves: Sta- tionary theory in Geometry and arithmetic , 283–307, EMS Ser. Congr. Rep., Eur. Math. Soc., Z¨ urich, 2012

2012
[22]

Pandharipande and A

R. Pandharipande and A. Pixton, Descendent theory for stable pairs on toric 3-folds , Jour. Math. Soc. Japan. 65 (2013), 1337–1372

2013
[24]

Pandharipande and A

R. Pandharipande and A. Pixton, Gromov-Witten/Pairs correspon- dence for the quintic , JAMS 30 (2017), 389–449

2017
[25]

Pandharipande and R

R. Pandharipande and R. P. Thomas, Curve counting via stable pairs in the derived category , Invent Math. 178 (2009), 407–447

2009
[26]

Pandharipande and R

R. Pandharipande and R. P. Thomas, The 3-fold vertex via stable pairs , Geom. Topol. 13 (2009), 1835–1876

2009
[27]

Pandharipande and R

R. Pandharipande and R. Thomas, 13/2 ways of counting curves in Moduli spaces , 282–333, London Math. Soc. Lecture Note Ser., 411, Cambridge Univ. Press, Cambridge, 2014

2014
[28]

Pandharipande and H.-H

R. Pandharipande and H.-H. Tseng, Higher genus Gromov-Witten the- ory of Hilb n(C2) and CohFTs associated to local curves , Forum of Math- ematics Pi 7 (2019)

2019
[29]

Pardon, Universally counting curves in Calabi-Yau threefolds , arXiv:2308.02948 (2023)

J. Pardon, Universally counting curves in Calabi-Yau threefolds , arXiv:2308.02948 (2023)

2023 arXiv
[30]

Schimpf, Stable pairs on local curves and Bethe roots , in preparation

M. Schimpf, Stable pairs on local curves and Bethe roots , in preparation
[31]

Teleman, The structure of 2D semisimple ﬁeld theories , Invent

C. Teleman, The structure of 2D semisimple ﬁeld theories , Invent. Math. 188 (2012), 525–588

2012
[32]

van Bree, Virasoro constraints for moduli spaces of sheaves on sur- faces, Forum of Mathematics Sigma 11 (2023), 1–35

D. van Bree, Virasoro constraints for moduli spaces of sheaves on sur- faces, Forum of Mathematics Sigma 11 (2023), 1–35. Department of Mathematics, ETH Z ¨urich Z¨urich, Switzerland. E-mail address : rahul@math.ethz.ch

2023

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.

[1] [23]

Pandharipande and A

R. Pandharipande and A. Pixton, Gromov-Witten/Pairs descendent correspondence for toric 3-folds , Geom. Topol. 18 (2014), 2747–2821

work page 2014

[2] [1]

Behrend and B

K. Behrend and B. Fantechi, The intrinsic normal cone, Invent. Math. 128 (1997), 45–88. Moduli of curves and moduli of sheaves 17

work page 1997

[3] [2]

Bojko, W

A. Bojko, W. Lim, and M. Moreira, Virasoro constraints for moduli of sheaves and vertex algebras , Invent. Math. 236 (2024), 387–476

work page 2024

[4] [3]

Bryan and T

J. Bryan and T. Graber, The crepant resolution conjecture , In: Alge- braic Geometry–Seattle 2005, Part 1, 23–42, Proc. Sympos. Pur e Math. 80, Amer. Math. Soc., Providence, RI, 2009

work page 2005

[5] [4]

Bryan and R

J. Bryan and R. Pandharipande, Local Gromov-Witten theory of curves, JAMS 21 (2008), 101–136

work page 2008

[6] [5]

Douady and J.-L

A. Douady and J.-L. Verdier, S´ eminaire de G´ eom´ etrie Analytique de l’Ecole Normale Sup´ erieure 1974/1975, Ast´ erisque36-37 (1976), Ex- pos´ es VI-IX

work page 1976

[7] [6]

Givental, Gromov-Witten invariants and quantization of quadratic Hamiltonians, Moscow Math

A. Givental, Gromov-Witten invariants and quantization of quadratic Hamiltonians, Moscow Math. J. 1 (2001), 487–518

work page 2001

[8] [7]

Le Potier, Faisceaux semi-stable et syst` emes coh´ erents, in Vector bun- dles in algebraic geometry (Durham, 1993) , LMS Lecture Note Ser

J. Le Potier, Faisceaux semi-stable et syst` emes coh´ erents, in Vector bun- dles in algebraic geometry (Durham, 1993) , LMS Lecture Note Ser. 208, 179–239, Cambridge Univ. Press: Cambridge, 1995

work page 1993