Virasoro constraints for Hodge integrals

Xin Wang

REVIEW 4 major objections 2 minor 26 references

Virasoro constraints for Hodge integrals

T0 review · 4 major / 2 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read A Virasoro algebra of differential operators is conjectured to annihilate the total Hodge potential of any smooth projective variety, with proofs in genus zero and one.

desk verdict The paper proves two genuinely new vanishing identities for Hodge integrals in genus 1 and higher; the stress test's counterexample to equation (36) is based on a misreading of the metric contraction and does not hold up. read the letter →

arxiv 2506.10033 v1 pith:FIT3Z43S submitted 2025-06-10 math-ph math.AGmath.MP

Xin Wang This is my paper

classification math-phmath.AGmath.MP MSC 14N3553D4514H1017B68

keywords HodgeintegralsVirasoroconstraintsGromov-WitteninvariantsmoduliofstablemapsbundletotaldescendantpotentialtautologicalrelationsquantizationquadraticHamiltonians

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 proposes that Hodge integrals—descendant invariants on moduli spaces of stable maps multiplied by Chern characters of the Hodge bundle, the bundle of holomorphic differentials on the domain curve—are annihilated by an explicit sequence of differential operators forming a Virasoro algebra, for any smooth projective target. It proves the conjecture in genus zero for all operators, in genus one for the first nontrivial operator at every Hodge insertion degree, and in higher genus for the first operator when the Hodge insertion degree is large enough. These constraints are universal differential equations mixing descendant Gromov-Witten variables with Hodge insertion parameters, so if the conjecture holds, Hodge integrals are governed by the same algebraic structure as ordinary descendant Gromov-Witten invariants. A semisimple-target version of the conjecture also follows from reconstruction theory.

What carries the argument

The central object is the total Hodge potential $D^E(t,s)=\exp(\sum_{g\geq 0}\hbar^{2g-2}F^E_g(t,s))$. The load-bearing identity is the reformulation of the Hodge-potential annihilation equations as $D^E(t,s)=\exp(\sum_{k\geq 1}\frac{B_{2k}}{(2k)!}s_k\,\widehat{z^{2k-1}})D(t)$, where $D(t)$ is the ordinary total descendant potential and the hat denotes the standard quantization of the quadratic Hamiltonian given by multiplication by $z^{2k-1}$. The Virasoro operators for Hodge integrals are the conjugates $L_n^E = \exp(\sum \cdots)\,L_n\,\exp(-\sum \cdots)$, so the difference operators $\Delta_{2k-1}$ applied to elementary symmetric functions of shifted degrees encode how this symplectic shift moves descendant indices. The proofs organize the genus expansion of $L_n^E D^E$ into the functions $\Psi^E_{g,n;k_1,\dots,k_m}$; vanishing of each function is a genus-$g$, degree-$(k_1,\dots,k_m)$ $L_n^E$-constraint.

What would settle it

Compute both sides of equation (36) for a concrete target, for example $X=\mathbb{P}^2$ with $\phi_\alpha$ the class of a point, using torus localization on the moduli space of stable maps. Any nonzero difference between $\langle\langle\tau_0(\phi_\alpha);\operatorname{ch}_1(E)\rangle\rangle_1$ and $\frac{1}{24}\sum_\beta \langle\langle\tau_0(\phi_\alpha)\phi_\beta\phi_\beta\rangle\rangle_0$ would refute Theorem 1.4 and Conjecture 1.1.

Watch

Extended reading notes

Core claim

Under the standing assumption $H^{\mathrm{odd}}(X;\mathbb{C})=0$, the paper defines operators $L_n^E$ for $n\geq -1$ on the big phase space of a smooth projective variety $X$ by conjugating the standard Virasoro operators $L_n$ with the exponential of a quadratic Hamiltonian built from Bernoulli numbers and the Hodge parameters $s_k$. It conjectures that $L_n^E D^E(t,s)=0$ for all $n$, and it proves the conjecture in three regimes: all $n$ in genus zero; the $L_1^E$ constraint in genus one with one Hodge character insertion for every $k_1\geq 1$; and the $L_1^E$ constraint for $g\geq 2$ when $k_1\geq \max\{g+1,(3g-1)/2\}$. The proofs expand $L_n^E D^E$ into functions $\Psi^E_{g,n;k_1,\dots,k_m}$ and show that each relevant function vanishes by the genus-zero Virasoro constraints, the annihilation equations for the Hodge potential, tautological relations on moduli of curves, and the vanishing of $\psi$-classes above the dimension of $M_{g,1}$ and $M_{g,2}$.

Load-bearing premise

The genus-one proof relies on the unproved claim in the paper's equation (36) that, for every smooth projective variety, a genus-one one-point Hodge insertion with $\operatorname{ch}_1(E)$ equals one twenty-fourth of the corresponding genus-zero three-point sum; if this equality fails, Theorem 1.4 and the conjecture fail.

Editorial extensions

If this is right

If the conjecture holds, Hodge integrals for any smooth projective variety satisfy universal linear differential equations that, together with the Hodge-potential annihilation equations, determine higher Hodge insertions from ordinary Gromov-Witten invariants.
The genus-zero theorem yields the $\widetilde{L}_n$ constraints for quantum cohomology of every target variety, recovering a family of differential equations previously proposed for Fano manifolds.
The genus-one theorem produces new universal identities relating genus-one Hodge integrals to genus-zero three-point functions for all targets, extending the known reduction of the genus-one Virasoro conjecture.
The higher-genus theorem gives new vanishing identities for pure Gromov-Witten invariants once Hodge insertions vanish by the rank bound, namely for $k_1\geq \max\{g+1,(3g-1)/2\}$.
For semisimple quantum cohomology, the conjecture holds in all genera, so Hodge potentials in that case are annihilated by all $L_n^E$ operators.

Reading between the lines

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

My inference: because the proofs use only genus-zero Virasoro constraints, the Hodge-potential annihilation equations, and tautological relations on moduli of curves, the same statements should extend to any compact symplectic target with $H^{\mathrm{odd}}=0$ and a well-defined Hodge bundle, not just smooth projective varieties.
My inference: the degree threshold in Theorem 1.5 is probably not optimal; the dimension vanishing that forces the final terms to vanish suggests that sharper bounds could be obtained by tracking the exact $\psi$-degree, and testing low-degree targets would reveal the correct cutoff.
My inference: the main obstruction to the full conjecture is not the Virasoro algebra structure but the availability of sufficiently many tautological relations on moduli of stable maps; if the needed relations hold in all genera, the same induction would likely prove Conjecture 1.1 completely.
My inference: proving equation (36) directly for a nontrivial target, for example by torus localization for a toric threefold, would be a sharp test of the genus-one conjecture, since that equation is currently the only unproved input in Theorem 1.4.

Share X Bluesky LinkedIn Reddit HN

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

Luke,

Quick take: this is a solid, workmanlike paper, not a breakthrough. It proves the genus-0 Virasoro constraints for Hodge integrals (essentially known from Liu–Yang–Zhang–Zhou, and the author says so in Remark 4.1), and then proves new vanishing identities in genus 1 and higher for the L_1^E operator: Theorem 1.4 for genus 1 with one Hodge insertion, and Theorem 1.5 for higher genus when the degree parameter is large enough. Those two are genuinely new to the literature, and they are proved with standard machinery: Faber–Pandharipande, genus-0 Virasoro, and the T-operator formalism of X. Liu.

I checked the stress-test note carefully. It claims equation (36) is false for X = P^1 at degree 0. That does not hold up. The sum on the right side of (36) is a metric contraction, i.e. the Casimir sum over the basis with the inverse intersection form. For P^1, with the standard basis {1, H}, the sum is 2 η^{12} ⟨1,1,H⟩ = 2, so the right side is 1/12, not 1/24. The stress test evaluated ⟨1,H,H⟩ as 1, but on P^1 the class H has degree 2 (real) and H^2 = 0, so that three-point function vanishes. The left side, from the virtual class on M_{1,1}(P^1,0), is indeed 1/12. The identity is a standard λ_1 = (1/12) δ_irr consequence and, with the correct contraction, it holds in this example.

So the load-bearing step in Lemma 5.8 is not obviously wrong. What is fair to say is that equation (36) is asserted without proof or reference, and it is load-bearing for Theorem 1.4. That is a real presentation gap, but a fixable one: a one-paragraph derivation from the boundary of M_{1,1} plus the splitting axiom would settle it. Similarly, the induction in Theorem 1.3 is sketched and the reduction in Theorem 1.5 is abbreviated. These are places where a referee should ask for more detail, not places where the paper collapses.

The paper is honest about what is new. The introduction and remarks correctly attribute the genus-0 result to [16]. No self-citation inflation, no invented entities. The algebra of the L_n^E operators is checked carefully, and the cancellations in the genus-1 computation are intricate and at least plausible.

Who benefits? People working on Hodge integrals, Virasoro constraints, and Gromov–Witten invariants of general targets. The paper is not for a general math-physicist. It deserves a serious referee; it is the kind of paper where the referee should be willing to spend time verifying the compressed steps. My recommendation: send it to peer review, with a note that the proof of equation (36) and the induction in Theorem 1.3 need expansion.

Best, Alex

Referee Report

4 major / 2 minor

Summary. The paper proposes Virasoro constraints for Hodge integrals in Gromov-Witten theory of arbitrary smooth projective varieties. It defines operators L_n^E as conjugates of the usual Virasoro operators by the Faber-Pandharipande exponential, and states Conjecture 1.1 that L_n^E D^E(t,s)=0. The main results are: Theorem 1.3, genus-zero constraints for all n; Theorem 1.4, genus-1, degree-k1 L_1^E constraints for k1≥1; and Theorem 1.5, higher-genus L_1^E constraints for k1≥max{g+1,(3g-1)/2}. The proofs use Givental quantization, the Faber-Pandharipande formula, genus-zero Virasoro constraints, tautological relations on M_{1,1}, and ψ-class dimension vanishings.

Significance. If the theorems are correct, the paper gives explicit universal differential equations for Hodge integrals of any target variety and a new hierarchy of constraints beyond the usual Virasoro conjecture. The explicit formula for L_n^E and the genus-zero theorem are useful and likely correct. The reduction of the higher-genus statements to ψ-class vanishings is a promising strategy. However, the proof of Theorem 1.4 rests on an asserted universal identity, equation (36), that appears to be false as written; this is a load-bearing issue. The paper also contains an arithmetic inconsistency in Lemma 5.3 and abbreviated arguments in Theorems 1.3 and 1.5.

major comments (4)

[§5.3, Lemma 5.3] The universal identity ⟨⟨τ_n(φ_α); ch_1(E)⟩⟩_1 = 1/24 ∑_β ⟨⟨τ_n(φ_α) φ_β φ_β⟩⟩_0 is asserted without proof and is not a direct consequence of the tautological relation λ_1=1/12 δ_irr on M_{1,1}. Pulling back δ_irr to M_{1,1}(X,A) involves the self-gluing locus in M_{0,2}(X,A) with ev_p=ev_q; the virtual class of this boundary carries an excess factor and the correct contraction should involve the metric on cohomology, not the unmeticized sum over β with the same class twice. A concrete check for X=P^1, A=0, n=0, φ_α=1 gives left side ∫_{M_{1,1}×P^1} λ_1 c_1(TP^1)=1/12, while the right side is ∑_β⟨1,φ_β,φ_β⟩_{0,0}=⟨1,1,1⟩+⟨1,H,H⟩=0+0=0. Thus Eq. (36) is false as stated, and Lemma 5.8, which is built on it, cannot support Theorem 1.4.
[§5.3, Lemma 5.3] In the proof of Lemma 5.3, after equations (21)–(22) the coefficient is computed as 2(2k1−1), but the displayed formula in the proof of Lemma 5.3 uses 2(2k1+1) in three places (e.g., the first lines of equation (24)). If 2(2k1+1) is the intended coefficient, the cancellation with Lemma 5.4 in equation (32) fails; if it is a typo, the computation must be corrected. This is a load-bearing arithmetic discrepancy in the proof of Theorem 1.4 for k1>1.
[§4.1, Theorem 1.3] The induction proving Ψ_{0,n;k1,...,km}=0 is only sketched. The displayed computation after the induction hypothesis is not fully derived: the action of \hat z^{2k_{m+1}-1}, the use of the induction hypothesis, and the role of the vanishing of ch_{2m-1}(E) on M_{0,n} are not explained in enough detail to verify. Since Theorem 1.3 is a central claim, the proof should be written out completely.
[§6.2, Theorem 1.5] The final step of the proof of Theorem 1.5 is too abbreviated. The text asserts that the left-hand side of equation (43) vanishes because ψ_1^m=0 for m>3g−2 on M_{g,1} and ψ_1^iψ_2^{m−i}=0 for m>3g−1 on M_{g,2}, but the preceding manipulations with the T and Q operators have not been shown to reduce the expression to such ψ-class integrals. Without this reduction, the conclusion is not justified.

minor comments (2)

[Throughout] There are numerous typographical errors, e.g., 'mainfold' in the introduction, 'Faber-Panharpande' in equation (5), 'hamitonians' and 'intergals' in Section 2.3, and the footnote 'H odd(X = 0)' should read 'H^{odd}(X)=0'.
[Abstract and §3.2] The term 'degree-(k1,...,km)' for the powers of the s-variables is potentially confusing because 'degree' is also used for the curve class A; consider renaming it 's-degree' or 'Hodge degree'.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: L_n^E is a transparent conjugation of known Virasoro operators; all load-bearing inputs are independent standard results.

full rationale

The derivation chain is not circular. The operator L_n^E is defined in Section 3.2 by the explicit conjugation L_n^E := exp(A)L_n exp(-A), with A = sum_k B_{2k}/(2k)! s_k [z^{2k-1}, and D^E(t,s) is related to D(t) by equation (5). The paper openly states the formal equivalence L_n^E D^E = 0 iff L_n D = 0; this is a definitional identity, not a hidden use of the conclusion. Theorem 1.3 is proved from the external genus-zero Virasoro constraints of Liu-Tian together with the standard vanishing ch_{2m-1}(E)=0 on M_{0,n}. The proof of Theorem 1.4 uses the Faber-Pandharipande differential equations, genus-zero Virasoro constraints, and the tautological relation lambda_1 = 1/12 delta_irr on M_{1,1}; the skeptic's objection to equation (36) is a correctness or validity challenge, not a circularity, since the asserted Hodge identity is an independent input rather than a restatement of the target vanishing. Theorem 1.5 similarly uses standard degree-dimension vanishing of psi classes and Chern character classes. No parameter is fitted, no prediction is a renamed fit, and no load-bearing step reduces by construction to its own input. The cited prior works [16], [17], [20], [21] are independent external results and are not by the present author, so there is no self-citation chain. The correct finding is therefore no significant circularity.

Assumptions & free parameters 0 free parameters · 8 assumptions · 0 invented entities

The paper introduces no fitted parameters and no new entities. It relies on standard theorems in Gromov-Witten theory, plus one asserted universal relation (36) that is not fully derived.

assumptions (8)

domain assumption H^odd(X;C)=0 for the target variety X
Stated in the introduction for simplicity; the paper says the general case follows with Z_2 grading (footnote in Section 3.2).
standard math Faber-Pandharipande formula D_{2l-1} D^E = 0 (equation (4)) and its equivalent quantization form (5)
Quoted from [7] and [9]; the paper does not reprove it.
standard math Givental quantization formalism, including the cocycle formula for commutators
Used throughout Section 3 to define L_n^E.
standard math Genus-0 Virasoro constraints for descendant Gromov-Witten invariants for any compact symplectic manifold (Liu-Tian)
Base case and key input for Theorem 1.3; cited as [21].
standard math Genus-1 topological recursion relation (25): ⟨τ_s(φ_β)⟩_1 = ∑_α ⟨τ_{s-1}(φ_β) φ_α⟩_0 ⟨φ_α⟩_1 + 1/24 ∑_α ⟨τ_{s-1}(φ_β) φ_α φ_α⟩_0
Used in Lemma 5.3 and 5.4; cited from [20].
domain assumption Universal equation (36) from the tautological relation λ_1 = 1/12 δ_irr on M_{1,1}
Asserted in Section 5.4 without detailed proof; load-bearing for Theorem 1.4.
standard math Mumford's vanishing ch_{2k-1}(E)=0 on M_{g,n} for k>g
Used in Section 6.2 to restrict the sum in Ψ^E_{g,1;k1}; cited to [22].
standard math Vanishing of ψ classes: ψ_1^m=0 for m>3g-2 on M_{g,1} and ψ_1^i ψ_2^j=0 for i+j>3g-1 on M_{g,2}
Used at the end of Section 6.2 to conclude vanishing in Theorem 1.5.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Virasoro constraints for Hodge integrals." pith.science (2026). https://pith.science/paper/FIT3Z43S

@misc{pith2026250610033,
  author       = {Pith},
  title        = {Pith review of: Virasoro constraints for Hodge integrals},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FIT3Z43S}},
  note         = {Machine review of arXiv:2506.10033}
}

abstract

The purpose of this paper is to study Virasoro constraints for Hodge integrals in Gromov-Witten theory of any target varieties. Results consist of the following: Firstly, we propose Virasoro conjecture for Hodge integrals in Gromov-Witten theory of any target varieties; Secondly, we prove Virasoro constraints for Hodge integrals in genus zero of any target varieties; Thirdly, we prove the genus-1 $L_1^{\mathbb{E}}$ constraint with one Hodge character class insertion for any target varieties; Lastly, we obtain certain new vanishing identities in higher genus for Gromov-Witten invariants of any target varieties.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

26 extracted references · 26 canonical work pages

[16]

S.-Q. Liu, D. Yang, Y. Zhang, and J. Zhou. The Virasoro-like algebra of a Frobenius manifold. Int. Math. Res. Not. IMRN , (16):13524–13561, 2023

work page 2023
[1]

Coates and A

T. Coates and A. Givental. Quantum Riemann-Roch, Lefschetz and Serre. Ann. of Math. (2), 165(1):15– 53, 2007

work page 2007
[2]

D. A. Cox and S. Katz. Mirror symmetry and algebraic geometry , volume 68 of Mathematical Surveys and Monographs. American Mathematical Society, Providence, RI, 1999

work page 1999
[3]

Dubrovin, S.-Q

B. Dubrovin, S.-Q. Liu, D. Yang, and Y. Zhang. Hodge integrals and tau-symmetric integrable hierar- chies of Hamiltonian evolutionary PDEs. Adv. Math., 293:382–435, 2016

work page 2016
[4]

Dubrovin and Y

B. Dubrovin and Y. Zhang. Frobenius manifolds and Virasoro constraints. Selecta Math. (N.S.) , 5(4):423–466, 1999

work page 1999
[5]

Eguchi, K

T. Eguchi, K. Hori, and C.-S. Xiong. Quantum cohomology and Virasoro algebra. Phys. Lett. B , 402(1- 2):71–80, 1997

work page 1997
[6]

Eguchi and C.-S

T. Eguchi and C.-S. Xiong. Quantum cohomology at higher genus: topological recursion relations and Virasoro conditions. Adv. Theor. Math. Phys. , 2(1):219–229, 1998

work page 1998
[7]

Faber and R

C. Faber and R. Pandharipande. Hodge integrals and Gromov-Witten theory.Invent. Math., 139(1):173– 199, 2000

work page 2000

Show all 26 references

[8]

E. Getzler. The Virasoro conjecture for Gromov-Witten invariants. In Algebraic geometry: Hirzebruch 70 (Warsaw, 1998) , volume 241 of Contemp. Math., pages 147–176. Amer. Math. Soc., Providence, RI, 1999

1998
[9]

A. B. Givental. Gromov-Witten invariants and quantization of quadratic Hamiltonians. volume 1, pages 551–568, 645. 2001. Dedicated to the memory of I. G. Petrovskii on the occasion of his 100th anniversary

2001
[10]

A. B. Givental. Semisimple Frobenius structures at higher genus. Internat. Math. Res. Notices , (23):1265–1286, 2001

2001
[11]

A. B. Givental. Symplectic geometry of Frobenius structures. In Frobenius manifolds, volume E36 of Aspects Math., pages 91–112. Friedr. Vieweg, Wiesbaden, 2004

2004
[12]

Guo and G

S. Guo and G. Wang. Virasoro constraints and polynomial recursion for the linear Hodge integrals. Lett. Math. Phys. , 107(4):757–791, 2017

2017
[13]

K. Hori. Constraints for topological strings in D ≥ 1. Nuclear Phys. B , 439(1-2):395–420, 1995

1995
[14]

Kontsevich

M. Kontsevich. Intersection theory on the moduli space of curves and the matrix Airy function. Comm. Math. Phys. , 147(1):1–23, 1992

1992
[15]

S.-Q. Liu, D. Yang, Y. Zhang, and C. Zhou. The loop equation for special cubic Hodge integrals. J. Differential Geom., 121(2):341–368, 2022. 34 XIN W ANG

2022
[17]

X. Liu. Elliptic Gromov-Witten invariants and Virasoro conjecture. Comm. Math. Phys. , 216(3):705– 728, 2001

2001
[18]

X. Liu. Idempotents on the big phase space. In Gromov-Witten theory of spin curves and orbifolds , volume 403 of Contemp. Math. , pages 43–66. Amer. Math. Soc., Providence, RI, 2006

2006
[19]

X. Liu. Genus-2 Gromov-Witten invariants for manifolds with semisimple quantum cohomology. Amer. J. Math. , 129(2):463–498, 2007

2007
[20]

X. Liu. On certain vanishing identities for Gromov-Witten invariants. Trans. Amer. Math. Soc. , 363(6):2939–2953, 2011

2011
[21]

Liu and G

X. Liu and G. Tian. Virasoro constraints for quantum cohomology. J. Differential Geom., 50(3):537–590, 1998

1998
[22]

D. Mumford. Towards an enumerative geometry of the moduli space of curves. In Arithmetic and geometry, Vol. II , volume 36 of Progr. Math., pages 271–328. Birkh¨ auser Boston, Boston, MA, 1983

1983
[23]

Okounkov and R

A. Okounkov and R. Pandharipande. Virasoro constraints for target curves. Invent. Math. , 163(1):47– 108, 2006

2006
[24]

C. Teleman. The structure of 2D semi-simple field theories. Invent. Math. , 188(3):525–588, 2012

2012
[25]

E. Witten. Two-dimensional gravity and intersection theory on moduli space. In Surveys in differential geometry (Cambridge, MA, 1990) , pages 243–310. Lehigh Univ., Bethlehem, PA, 1991

1990
[26]

J. Zhou. On recursion relation for hodge integrals from the cut-and-join equations. preprint, 2011. Department of Mathematics, Shandong University, Jinan, China Email address : xinwmath@gmail.com

2011

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [16]

S.-Q. Liu, D. Yang, Y. Zhang, and J. Zhou. The Virasoro-like algebra of a Frobenius manifold. Int. Math. Res. Not. IMRN , (16):13524–13561, 2023

work page 2023

[2] [1]

Coates and A

T. Coates and A. Givental. Quantum Riemann-Roch, Lefschetz and Serre. Ann. of Math. (2), 165(1):15– 53, 2007

work page 2007

[3] [2]

D. A. Cox and S. Katz. Mirror symmetry and algebraic geometry , volume 68 of Mathematical Surveys and Monographs. American Mathematical Society, Providence, RI, 1999

work page 1999

[4] [3]

Dubrovin, S.-Q

B. Dubrovin, S.-Q. Liu, D. Yang, and Y. Zhang. Hodge integrals and tau-symmetric integrable hierar- chies of Hamiltonian evolutionary PDEs. Adv. Math., 293:382–435, 2016

work page 2016

[5] [4]

Dubrovin and Y

B. Dubrovin and Y. Zhang. Frobenius manifolds and Virasoro constraints. Selecta Math. (N.S.) , 5(4):423–466, 1999

work page 1999

[6] [5]

Eguchi, K

T. Eguchi, K. Hori, and C.-S. Xiong. Quantum cohomology and Virasoro algebra. Phys. Lett. B , 402(1- 2):71–80, 1997

work page 1997

[7] [6]

Eguchi and C.-S

T. Eguchi and C.-S. Xiong. Quantum cohomology at higher genus: topological recursion relations and Virasoro conditions. Adv. Theor. Math. Phys. , 2(1):219–229, 1998

work page 1998

[8] [7]

Faber and R

C. Faber and R. Pandharipande. Hodge integrals and Gromov-Witten theory.Invent. Math., 139(1):173– 199, 2000

work page 2000