betaFeedback

WorksMathematics

Quantum Cohomology and irrationality of Küchle fourfolds

Abstract

We prove that very general Küchle fourfolds of types (c1), (c6), (b8), and (b11) are irrational. The key ingredients are a new rationality criterion, based on the theory of Hodge atoms due to Katzarkov–Kontsevich–Pantev–Yu and its integral extension by Guéré, and a computation of the action of quantum multiplication on the vanishing cohomology of these fourfolds, which uses a version of the Quantum Lefschetz Hyperplane Principle. Quantum multiplication acts as a scalar on the vanishing cohomology, and we give an enumerative interpretation of this scalar.

Introduction

In [Küc95], Küchle classified all smooth index-one Fano fourfolds that arise as the zero locus of a regular section of a globally generated, completely reducible homogeneous vector bundle on a Grassmannian. Our main result concerns their rationality.

Theorem 1.1. Very general Küchle fourfolds of types (b8), (b11), (c1), and (c6) are irrational.

What makes the birational geometry of these fourfolds particularly compelling is that their middle cohomology carries a large surface-type Hodge structure that is not of K3 type. Whereas the obstructions to rationality for cubic, Gushel–Mukai, and Verra fourfolds [Gué26b, BMP26, Gué26a] involve K3 surfaces, rationality of a very general fourfold of type (c1) or (c6) would imply the existence of a projective surface SS of geometric genus 5 or 2, respectively.

The proof of Theorem 1.1 rests on a rationality criterion for Fano fourfolds, which we now state. Let XX be a smooth projective complex fourfold and define

ΛX:=H4(X,Z)/tors,AX:=ΛX∩H2,2(X),TX:=AX⊥⊆ΛX,\Lambda_X := H^4(X,\mathbb{Z})/\mathrm{tors}, \qquad A_X := \Lambda_X \cap H^{2,2}(X), \qquad T_X := A_X^\perp\subseteq\Lambda_X,

where orthogonality is taken with respect to the intersection pairing (α,β)X=∫Xα⌣β(\alpha,\beta)_X = \int_X \alpha\smile\beta; we write TX,Q:=TX⊗ZQT_{X,\mathbb{Q}} := T_X \otimes_{\mathbb{Z}} \mathbb{Q}. For a smooth projective surface SS, write

ΛS:=H2(S,Z)/tors,TS:=(NS(S)/tors)⊥⊂ΛS,ρ(S):=rk⁡(NS(S)/tors)\Lambda_S := H^2(S,\mathbb{Z})/\mathrm{tors}, \qquad T_S := (\mathrm{NS}(S)/\mathrm{tors})^\perp\subset\Lambda_S, \qquad\rho(S) := \operatorname{rk}(\mathrm{NS}(S)/\mathrm{tors})

We use the following notation for a nondegenerate lattice LL, as recalled in Section 5:

D(L)D(L)the discriminant group L∨/LL^\vee/L
ℓ(D(L))\ell(D(L))the minimal number of generators of D(L)D(L)

As we will describe in Section 4.1 below, the cohomology group H∙(X;K)H^\bullet(X;K), where KK is a suitable coefficient field, carries an endomorphism UX(τ)U_X(\tau) given by quantum multiplication by the Euler vector field, depending on parameters τ\tau in the direction of rational Hodge classes. If XX satisfies (1.1) below, then UX(τ)U_X(\tau) acts on the transcendental cohomology TX⊗ZKT_X \otimes_{\mathbb{Z}} K by a scalar, as we show in the proof of Proposition 1.2: let us call this scalar λX(τ)\lambda_X(\tau). Put λX0=λX(0)\lambda_X^0=\lambda_X(0), and define, for general τ\tau,

mX:=mult⁡λX(τ)det⁡(t id−UX(τ)),MX:=mult⁡λX0det⁡(t id−UX(0)).m_X := \operatorname{mult}_{\lambda_X(\tau)}\det(t\,\mathrm{id}-U_X(\tau)), \qquad M_X := \operatorname{mult}_{\lambda_X^0}\det(t\,\mathrm{id}-U_X(0)).

Here mXm_X is the dimension of the generalized eigenspace of UX(τ)U_X(\tau) containing TXT_X, while MXM_X is the specialization of this quantity to the small quantum locus τ=0\tau=0. Since multiplicities can only jump up at special points, mX≤MXm_X \le M_X.

Proposition 1.2 (Quantum Noether integral obstruction). Let XX be a smooth projective complex fourfold satisfying

h2,0(X)=h1,0(X)=0,h3,1(X)>0,End⁡Hdg(TX,Q)=Q id.(1)h^{2,0}(X)=h^{1,0}(X)=0,\qquad h^{3,1}(X)>0,\qquad\operatorname{End}_{\mathrm{Hdg}}(T_X,\mathbb{Q})=\mathbb{Q}\,\mathrm{id}. \tag*{(1)}

Let r=rk⁡TXr=\operatorname{rk}T_X. If XX is rational, there is a smooth minimal projective surface SS and an integral Hodge isometry TX(1)≃TST_X(1)\simeq T_S such that:

(1) We have

dim⁡QH∙(S,Q)≤mX≤MX\dim_{\mathbb{Q}} H^\bullet(S,\mathbb{Q}) \le m_X \le M_X

and in particular, r+max⁡{1,ℓ(D(TX))}+2≤mX≤MXr+\max\{1,\ell(D(T_X))\}+2\le m_X\le M_X.

(2) The Picard-rank ρ(S)\rho(S) of SS satisfies

ℓ(D(TX))≤ρ(S)≤10+12h3,1(X)−r.\ell(D(T_X))\le\rho(S)\le10+12h^{3,1}(X)-r.

(3) If MX≤r+3M_X\le r+3, then

ρ(S)=1,h1,0(S)=0,pg(S)=h3,1(X),\rho(S)=1,\qquad h^{1,0}(S)=0,\qquad p_g(S)=h^{3,1}(X),

the group D(TX)D(T_X) is cyclic, and

∣D(TX)∣k2=9+12h3,1(X)−rfor some k∈Z.(2)|D(T_X)|k^2=9+12h^{3,1}(X)-r \qquad\text{for some } k\in\mathbb{Z}. \tag*{(2)}

The criterion also recovers the recent result of Guéré:

Corollary 1.3 (Guéré, [Gué26a, Corollary 28]). A very general Verra fourfold is irrational.

Kŭchle Typeh3,1(X)h^{3,1}(X)rk⁡TX\operatorname{rk}T_XMXM_XD(TX)D(T_X)Rationality Criterion
(b8)34245Z/2⊕Z/6\mathbb{Z}/2\oplus\mathbb{Z}/62≤ρ(S)≤12\le\rho(S)\le1
(b11)23235(Z/3)2(\mathbb{Z}/3)^22≤ρ(S)≤12\le\rho(S)\le1
(c1)56063Z/21\mathbb{Z}/2121k2=921k^2=9
(c6)23235Z/9\mathbb{Z}/99k2=19k^2=1

Table 1. Rationality obstructions for very general Kŭchle fourfolds obtained from Proposition 1.2.

Related work. Amongst the twenty-one families of Küchle varieties [Küc95], only those of types (c5), (c7), and (d3) are of K3 type [4 Section 2.3]. Kuznetsov studied the geometry of the varieties of types (b4), (b9), (c7), and (d3) in [30], where he proved type (b9) to be rational [30 Theorem 2.5].

Kuznetsov also studied the birational geometry of the Küchle fivefold of type (c5) in [31]. Taking its smooth hyperplane section gives the Küchle fourfold of type (c5), whose very general member was proven to be irrational by Benedetti–Fay–Guéré–Manivel–Perrin [3 Proposition 5.1]. This result uses Noether’s formula to bound the rank of the transcendental part of a surface atom [3 Theorem 4.1]; here we combine it with a dimension bound from small quantum multiplication and integral lattice data. The type (c5) irrationality result was recovered by Alexeev–Cavenaghi–Galindo–Georgiev–Katzarkov–Martins in [2], who also prove irrationality of very general Küchle fourfolds of types (c2) and (b10) [2 Theorem 6.13 and 6.16]. Both papers have related monodromy arguments concerning the scalar action on vanishing cohomolology; see [3 proof of Theorem 4.1] and [2 Theorem 5.35].

For the fourfold of type (c7), Kuznetsov identified it with the blowup of a cubic fourfold along a Veronese surface in [30 Corollary 4.11]. Using this description, the author proved a very general Küchle fourfold of type (c7) to be irrational [15 Corollary 1.4].

The motivating thread throughout this paper’s work on birational geometry is the work of Katzarkov–Kontsevich–Pantev–Yu in [27]. To apply it to the fourfolds considered here, we use Guéré’s recent extension to integral Hodge structures [Gué26a, Proposition 26]. Our approach for understanding and recovering the relevant quantum cohomology is driven by the Quantum Lefschetz and Riemann–Roch methods of Coates–Givental [10, 7]. Our motivation for applying it within the context of Fano fourfolds finds itself in the large number of quantum periods computed by Coates–Galkin–Kasprzyk–Strangeway in [8].

Outline

Outline. In Section 2 we show that quantum multiplication by the hyperplane class HH acts by a scalar on vanishing cohomology, and that this scalar is a repeated eigenvalue on the inherited part of cohomology with a nonzero isotropic eigenvector. In Section 3, we express this scalar as a line count and construct embeddings of vanishing and transcendental cohomology using the Fano surface of lines. We recall the Quantum Lefschetz theorem in Section 4 and apply it to the Küchle fourfolds of type (c1), (c6), (b8), and (b11). In Section 5 we prove our rationality criterion, which we apply to these Küchle fourfolds in Section 6. We also apply it to Verra fourfolds in Section 7. All varieties throughout this paper are smooth and projective over C\mathbb{C}.

Monodromy and Vanishing Cohomology

Let X⊂Y⊂PNX \subset Y \subset\mathbb{P}^{N} be a smooth hyperplane section of a smooth projective fivefold, and let i:X↪Yi:X \hookrightarrow Y denote the inclusion. Assume that XX is an index-one Fano fourfold of Picard-rank one with Hodd(X,Q)=0H^{\mathrm{odd}}(X,\mathbb{Q})=0. Write H=c1(X)H=c_1(X) and set K=Q(q)K=\mathbb{Q}(q).

Define the inherited and vanishing cohomology by

BX:=im⁡(i∗:H∙(Y,Q)→H∙(X,Q)),VX:=(BX∩H4(X,Q))⊥⊂H4(X,Q),B_X := \operatorname{im}(i^*:H^\bullet(Y,\mathbb{Q})\to H^\bullet(X,\mathbb{Q})), \qquad V_X := \left(B_X\cap H^4(X,\mathbb{Q})\right)^\perp\subset H^4(X,\mathbb{Q}),

where orthogonal complements are taken with respect to the Poincaré pairing. By weak and hard Lefschetz, the pairing on BXB_X is nondegenerate and we have the orthogonal decomposition

H∙(X,Q)=Heven(X,Q)=BX⊕VX.H^\bullet(X,\mathbb{Q})=H^{\mathrm{even}}(X,\mathbb{Q})=B_X\oplus V_X.

Let U⊂∣OY(1)∣U \subset|{\cal O}_Y(1)| be the locus of smooth hyperplane sections and u0=[X]∈Uu_0=[X]\in U. The local system of vanishing cohomology over UU defines the monodromy representation

ρvan:π1(U,u0)⟶O(VX),Γ:=im⁡(ρvan),\rho_{\mathrm{van}}:\pi_1(U,u_0)\longrightarrow O(V_X),\qquad\Gamma:=\operatorname{im}(\rho_{\mathrm{van}}),

where Γ\Gamma is the monodromy group; see [11 §5].

Gromov–Witten invariants

We briefly recall stable maps and Gromov–Witten invariants; see [28, 35]. For a smooth projective variety ZZ and a class β∈H2(Z,Z)\beta\in H_2(Z,\mathbb{Z}), let M‾0,n(Z,β)\overline{\mathcal{M}}_{0,n}(Z,\beta) be the moduli space of stable maps

f:(C,p1,…,pn)⟶Zwithf∗[C]=β.f:(C,p_1,\ldots,p_n)\longrightarrow Z\quad\text{with}\quad f_*[C]=\beta.

Here CC is a connected projective nodal curve of arithmetic genus-zero with distinct smooth marked points p1,…,pn∈Cp_1,\ldots,p_n\in C, and every irreducible component of CC contracted by ff contains at least three special points (marked points or nodes).

For each jj, there is an evaluation map

ev⁡j:M‾0,n(Z,β)⟶Z,ev⁡j(f)=f(pj).\operatorname{ev}_j:\overline{\mathcal{M}}_{0,n}(Z,\beta)\longrightarrow Z,\qquad\operatorname{ev}_j(f)=f(p_j).

The space M‾0,n(Z,β)\overline{\mathcal{M}}_{0,n}(Z,\beta) carries a virtual fundamental class [M‾0,n(Z,β)]vir[\overline{\mathcal{M}}_{0,n}(Z,\beta)]^{\mathrm{vir}} of expected dimension dim⁡Z−3+n+c1(Z)⋅β\dim Z-3+n+c_1(Z)\cdot\beta.

The genus-zero Gromov–Witten invariants of ZZ are then the rational virtual intersection numbers

⟨α1ψ1k1,…,αnψnkn⟩0,n,βZ:=∫[M‾0,n(Z,β)]vir∏j=1nev⁡j∗αj⌣ψjkj,αj∈H∙(Z,Q),\left\langle\alpha_1\psi_1^{k_1},\ldots,\alpha_n\psi_n^{k_n}\right\rangle^Z_{0,n,\beta}:=\int_{[\overline{\mathcal{M}}_{0,n}(Z,\beta)]^{\mathrm{vir}}}\prod_{j=1}^{n}\operatorname{ev}_j^*\alpha_j\smile\psi_j^{k_j},\qquad\alpha_j\in H^\bullet(Z,\mathbb{Q}),

where ψj\psi_j is the first Chern class of the line bundle whose fibre at ff is the cotangent line of CC at pjp_j; until Section 4 all kjk_j are zero. Note that these invariants do not change when ZZ is deformed, and they satisfy the divisor equation

⟨D,α1,…,αn⟩0,n+1,βZ=(D⋅β)⟨α1,…,αn⟩0,n,βZ\left\langle D,\alpha_1,\ldots,\alpha_n\right\rangle^Z_{0,n+1,\beta}=(D\cdot\beta)\left\langle\alpha_1,\ldots,\alpha_n\right\rangle^Z_{0,n,\beta}

for D∈H2(Z,Q)D\in H^2(Z,\mathbb{Q}) and β≠0\beta\ne0. When ZZ carries a hyperplane class HH, we write an integer dd in place of β\beta for the disjoint union of moduli spaces, or the sum of the invariants, over all β∈H2(Z,Z)\beta\in H_2(Z,\mathbb{Z}) with H⋅β=dH\cdot\beta=d.

The quantum product

For the fourfold XX with H=c1(X)H=c_1(X), the small quantum product on H∙(X,K)H^\bullet(X,K) is the KK-bilinear product defined by

(a⋆b,c)=∑d≥0qd⟨a,b,c⟩0,3,dXfor all a,b,c∈H∙(X,Q).(a\star b,c)=\sum_{d\geq0}q^d\left\langle a,b,c\right\rangle^X_{0,3,d}\qquad\text{for all }a,b,c\in H^\bullet(X,\mathbb{Q}).

Here the Novikov variable qq records the degree and the term with d=0d=0 is ∫Xa⌣b⌣c\int_X a\smile b\smile c. By the dimension formula, the coefficient of qdq^d in a⋆ba\star b has degree deg⁡a+deg⁡b−2d\deg a+\deg b-2d, so the sum is finite. The product ⋆\star is graded commutative, associative, and has unit 1∈H0(X,K)1\in H^0(X,K).

Write

B:=BX⊗QK,V:=VX⊗QK.B:=B_X\otimes_{\mathbb{Q}}K,\qquad V:=V_X\otimes_{\mathbb{Q}}K.

The small quantum product ⋆\star makes

R:=H∙(X,K)=B⊕VR:=H^\bullet(X,K)=B\oplus V

a Frobenius algebra: that is, a finite-dimensional unital commutative associative KK-algebra equipped with a nondegenerate symmetric bilinear pairing satisfying (a⋆b,c)=(a,b⋆c)(a\star b,c)=(a,b\star c).

This compatibility, together with the monodromy equivariance of ⋆\star proved in Proposition 2.1, leads to Proposition 2.2: an abstract Frobenius algebra version of Hu’s eigenvector construction for projective complete intersections; see [20 Proposition 5.1].

Proposition 2.1. If the pairing on VX⊗QRV_X \otimes_{\mathbb{Q}} \mathbb{R} is indefinite, then Γ\Gamma is Zariski dense in the orthogonal group O(VX)O(V_X). Consequently, the small quantum product on RR is O(V)O(V)-equivariant—with O(V)O(V) acting trivially on BB and by its standard representation on VV.

Proof. By standard theory, Γ\Gamma is generated by Picard–Lefschetz reflections in the vanishing cycles of a Lefschetz pencil through XX. These cycles span VXV_X and form a single orbit up to sign. Thus VXV_X is an absolutely irreducible Γ\Gamma-module.

By Deligne’s reflection-group alternative [12 Lemme 4.4.2s] (see also [26 Part II] for a statement in English), the Zariski closure of Γ\Gamma is either a finite reflection group or O(VX)O(V_X). It cannot be finite, as assuming so contradicts the stated indefiniteness: a finite group preserves a positive definite form on VX⊗RRV_X \otimes_{\mathbb{R}} \mathbb{R}, which by absolute irreducibility and Schur’s Lemma is proportional to the pairing, which would then be definite.

Finally, deformation invariance of Gromov–Witten invariants makes the small quantum product equivariant under Γ\Gamma. This is a Zariski-closed condition, and hence extends to O(VX)O(V_X) and, after extending scalars, to O(V)O(V).

Proposition 2.2. Let R=B⊕VR = B \oplus V be an orthogonal decomposition of a Frobenius algebra over a field kk of characteristic zero, with 1∈B1 \in B and dim⁡kV≥2\dim_k V \ge2, whose product ⋆\star is O(V)O(V)-equivariant as in Proposition 2.1.

Then BB is a subalgebra, and there are a unique character χ:B→k\chi: B \to k and a unique class bχ∈Bb_\chi\in B such that

a⋆v=χ(a)v,v⋆w=(v,w)bχ,(bχ,a)=χ(a)(3)a \star v = \chi(a) v,\qquad v \star w = (v,w)b_\chi,\qquad(b_\chi,a) = \chi(a) \tag*{(3)}

for all a∈Ba \in B and v,w∈Vv,w \in V; in particular (bχ,1)=1(b_\chi,1) = 1. Moreover,

a⋆bχ=χ(a)bχandχ(bχ)=0,(4)a \star b_\chi= \chi(a)b_\chi\qquad\text{and} \qquad\chi(b_\chi) = 0, \tag*{(4)}

so that bχb_\chi is isotropic and bχ⋆bχ=0b_\chi\star b_\chi= 0. Finally, for H∈BH \in B, writing LB:=(H⋆)∣BL_B := (H\star)|_B and λ:=χ(H)\lambda:= \chi(H),

(t−λ)2 divides det⁡(t id⁡B−LB).(5)(t-\lambda)^2 \text{ divides } \det(t\,\operatorname{id}_B-L_B). \tag*{(5)}

Proof. Equivariance of the quantum product under id⁡B⊕(−id⁡V)\operatorname{id}_B \oplus(-\operatorname{id}_V) gives

B⋆B⊆B,B⋆V⊆V,V⋆V⊆B.B \star B \subseteq B,\qquad B \star V \subseteq V,\qquad V \star V \subseteq B.

The standard representation of O(V)O(V) is absolutely irreducible, so Schur’s lemma gives a⋆v=χ(a)va \star v = \chi(a)v. Since ⋆\star is associative and unital, χ\chi is an algebra homomorphism. Nondegeneracy of the pairing on BB provides a unique bχ∈Bb_\chi\in B with (bχ,a)=χ(a)(b_\chi,a) = \chi(a) for all a∈Ba \in B. For v,w∈Vv,w \in V and a∈Ba \in B, the Frobenius property then gives

(v⋆w,a)=(v,w⋆a)=χ(a)(v,w)=((v,w)bχ,a),(v \star w,a) = (v,w \star a) = \chi(a)(v,w) = ((v,w)b_\chi,a),

so v⋆w=(v,w)bχv \star w = (v,w)b_\chi, again by nondegeneracy. This proves (2.1).

Now fix v,w∈Vv,w \in V with (v,w)=1(v,w) = 1, so that bχ=v⋆wb_\chi= v \star w. For a∈Ba \in B, associativity gives a⋆bχ=(a⋆v)⋆w=χ(a)bχa \star b_\chi= (a \star v) \star w = \chi(a)b_\chi. Since dim⁡V≥2\dim V \ge2, there is a nonzero u∈Vu \in V orthogonal to ww; then w⋆u=0w \star u = 0, so

χ(bχ)u=bχ⋆u=(v⋆w)⋆u=v⋆(w⋆u)=0\chi(b_\chi)u = b_\chi\star u = (v \star w) \star u = v \star(w \star u) = 0

and therefore χ(bχ)=0\chi(b_\chi) = 0. This proves (2.2), whence (bχ,bχ)=χ(bχ)=0(b_\chi,b_\chi) = \chi(b_\chi) = 0 and $b_\chi \star b_\chi = \chi(b_\chi)b_\chi = 0.

Finally, bχ≠0b_{\chi} \ne0 because (bχ,1)=χ(1)=1(b_{\chi},1)=\chi(1)=1, so kbχk b_{\chi} is a line in the hyperplane ker⁡χ\ker\chi. Both are preserved by LBL_B, which acts by λ\lambda on kbχk b_{\chi} and on B/ker⁡χB/\ker\chi, since H⋆bχ=λbχH \star b_{\chi}=\lambda b_{\chi} and χ(H⋆a)=λχ(a)\chi(H \star a)=\lambda\chi(a) for a∈Ba \in B. This gives (2.3).

When det⁡(t−LB)\det(t-L_B) has more than one repeated root, as for types (b11) and (c1) below, (2.1) and (2.2) decide between them, as the eigenspace of λ\lambda contains the isotropic class bχb_{\chi}. For an example of this, see the proof of Proposition 4.6.

Lines and the Cylinder Correspondence

Keeping the notation of the previous section, we identify the scalar λ\lambda as −q-q times the number of lines of YY through a general point. Although we do not use this result in the proof of Theorem 1.1, where λ\lambda is found from (2.2) and (2.3), it provides a geometric interpretation of its value. By a line we mean a curve of degree one in PN\mathbb{P}^{N}, where degrees of stable maps to XX and YY are taken with respect to HH.

Proposition 3.1 (Lines and the vanishing eigenvalue). Let X⊂YX \subset Y be a smooth hyperplane section of a smooth projective fivefold with −KY=2H-K_Y=2H, where HH is the hyperplane class. Suppose that the Fano schemes of lines F(Y)F(Y) and F(X)F(X) are smooth of dimensions 4 and 2, and let aY=⟨[pt]⟩0,1,1Ya_Y=\langle[\mathrm{pt}]\rangle^{Y}_{0,1,1} be the number of lines of YY through a general point. Then

⟨v,w⟩0,2,1X=−aY(v,w)Xfor all v∈VX and w∈H4(X,Q).(6)\langle v,w\rangle^{X}_{0,2,1}=-a_Y(v,w)_X \qquad\text{for all } v\in V_X \text{ and } w\in H^4(X,\mathbb{Q}). \tag*{(6)}

In particular, if H⋆H \star acts on VV by a scalar, then this scalar is −aYq-a_Yq.

Proof. Since HH is very ample, a degree one stable map with two marked points is a line ℓ⊂Y\ell\subset Y with p1,p2∈ℓp_1,p_2 \in\ell. When these points collide, they move into a new component or “bubble”, attached to the line at a node, and contracted by the map to a single point of YY. In either case,

MY:=M‾0,2(Y,1)=PY×F(Y)PY,MX:=M‾0,2(X,1)=PX×F(X)PX.M_Y:=\overline{M}_{0,2}(Y,1)=P_Y \times_{F(Y)} P_Y,\qquad M_X:=\overline{M}_{0,2}(X,1)=P_X \times_{F(X)} P_X.

where P→FP \to F denotes the universal line. The stable maps with a bubble form the diagonal DD of MYM_Y, which is a copy of PY=M‾0,1(Y,1)P_Y=\overline{M}_{0,1}(Y,1). By assumption MYM_Y, MXM_X and DD are smooth of dimensions 6, 4 and 5, which are the expected dimensions since −KY=2H-K_Y=2H and −KX=H-K_X=H. In particular, the evaluation map ev⁡D:D→Y\operatorname{ev}_D:D \to Y has degree aYa_Y.

Let ev⁡=(ev⁡1,ev⁡2)\operatorname{ev}=(\operatorname{ev}_1,\operatorname{ev}_2) denote the evaluation maps and put

IY:=ev⁡∗[MY]∈H8(Y×Y,Q),IX:=ev⁡∗[MX]∈H8(X×X,Q).I_Y:=\operatorname{ev}_*[M_Y]\in H^8(Y\times Y,\mathbb{Q}),\qquad I_X:=\operatorname{ev}_*[M_X]\in H^8(X\times X,\mathbb{Q}).

We claim that

(i×i)∗IY=IX+aY[ΔX]in H8(X×X,Q),(7)(i\times i)^*I_Y=I_X+a_Y[\Delta_X]\qquad\text{in } H^8(X\times X,\mathbb{Q}), \tag*{(7)}

where ΔX\Delta_X is the diagonal.

A line in YY either lies in XX or meets XX in a single point. Hence

W:=ev⁡−1(X×X)=MX∪DX,DX:=ev⁡D−1(X)⊂D,W:=\operatorname{ev}^{-1}(X\times X)=M_X\cup D_X,\qquad D_X:=\operatorname{ev}_D^{-1}(X)\subset D,

and both pieces have dimension 4=dim⁡MY−24=\dim M_Y-2. Each occurs with multiplicity one. Indeed, if s∈H0(Y,OY(1))s\in H^0(Y,\mathcal{O}_Y(1)) defines XX, then WW is cut out by s(p1)=s(p2)=0s(p_1)=s(p_2)=0. Where p1≠p2p_1\ne p_2, these equations say that ss vanishes identically on ℓ\ell, so they cut out MXM_X, which is smooth of codimension two.

Near a line ℓ⊄X\ell\not\subset X, write σ(ℓ′)=ℓ′∩X\sigma(\ell')=\ell'\cap X for nearby lines ℓ′\ell'. Then p1=σ(ℓ′)p_1=\sigma(\ell') and p2=σ(ℓ′)p_2=\sigma(\ell'), which again cut out a smooth subvariety of codimension two. The remaining locus consists of the bubbles on lines contained in XX and has dimension three. Thus

(i×i)′[MY]=[MX]+[DX],(i\times i)'[M_Y]=[M_X]+[D_X],

by [16 Proposition 7.1]. Pushing forward by [16 Theorem 6.2(a)] gives (7): the map DX→X×XD_X\to X\times X factors through the diagonal ΔX≅X\Delta_X\cong X and the induced map ev⁡DX:DX→X\operatorname{ev}_{D_X}:D_X\to X satisfies

(ev⁡DX)∗[DX]=aY[X],(\operatorname{ev}_{D_X})_*[D_X]=a_Y[X],

by compatibility of proper pushforward with the refined pullback along ii.

Now let v∈VXv\in V_X and w∈H4(X,Q)w\in H^4(X,\mathbb{Q}). Since VXV_X is orthogonal to i∗H4(Y,Q)i^*H^4(Y,\mathbb{Q}), we have i∗v=0i_*v=0, and so

∫X×X(v⊠w)⌣(i×i)∗IY=∫Y×Y(i∗v⊠i∗w)⌣IY=0.\int_{X\times X}(v\boxtimes w)\smile(i\times i)^*I_Y=\int_{Y\times Y}(i_*v\boxtimes i_*w)\smile I_Y=0.

By (7) this reads ⟨v,w⟩0,2,1X+aY(v,w)X=0\langle v,w\rangle^{X}_{0,2,1}+a_Y(v,w)_X=0, which proves (6): on vanishing classes, the contribution of the lines of XX cancels that of the bubbles.

Finally, the coefficient of qdq^d in H⋆vH\star v lies in H6−2d(X,Q)H^{6-2d}(X,\mathbb{Q}), and by the divisor equation the coefficient of qq is

∑j⟨v,ϕj⟩0,2,1Xϕj=−aYv,\sum_j\langle v,\phi_j\rangle^{X}_{0,2,1}\phi^j=-a_Yv,

where {ϕj}\{\phi_j\} and {ϕj}\{\phi^j\} are dual bases of H4(X,Q)H^4(X,\mathbb{Q}). So if H⋆v=λvH\star v=\lambda v, then λ=−aYq\lambda=-a_Yq.

Gromov–Witten invariants and the decomposition BX⊕VXB_X\oplus V_X are invariant under deformations of the pair (Y,X)(Y,X). So the conclusions of Proposition 3.1 hold as soon as (Y,X)(Y,X) can be deformed to a pair for which F(Y)F(Y) and F(X)F(X) are smooth of the expected dimensions.

Corollary 3.2 (The cylinder correspondence). In the situation of Proposition 3.1, let p:P→Fp:P\to F be the universal line over F=F(X)F=F(X) and e:P→Xe:P\to X its evaluation map, and let

Φ=p∗e∗:H4(X,Q)→H2(F,Q)\Phi=p_*e^*:H^4(X,\mathbb{Q})\to H^2(F,\mathbb{Q})

be the Abel–Jacobi map.

(Φv,Φw)F=−aY(v,w)Xfor all v∈VX and w∈H4(X,Q).(8)(\Phi v,\Phi w)_F=-a_Y(v,w)_X \qquad\text{for all }v\in V_X\text{ and }w\in H^4(X,\mathbb{Q}). \tag*{(8)}

In particular, if aY≠0a_Y\ne0, then Φ\Phi restricts to embeddings

Φ:VX(1)↪H2(F,Q),Φ:TX,Q(1)↪TF,Q,\Phi:V_X(1)\hookrightarrow H^2(F,\mathbb{Q}),\qquad\Phi:T_{X,\mathbb{Q}}(1)\hookrightarrow T_{F,\mathbb{Q}},

the second provided that BX∩H4(X,Q)B_X\cap H^4(X,\mathbb{Q}) consists of algebraic classes, so that TX,Q⊆VXT_{X,\mathbb{Q}}\subseteq V_X.

Proof. Since M‾0,2(X,1)=P×FP\overline{M}_{0,2}(X,1)=P\times_F P, the projection formula gives ⟨v,w⟩0,2,1X=(Φv,Φw)F\langle v,w\rangle^X_{0,2,1}=(\Phi v,\Phi w)_F for all v,w∈H4(X,Q)v,w\in H^4(X,\mathbb{Q}), so (8) is (6). It follows that Φ\Phi is injective on VXV_X when aY≠0a_Y\ne0, and Φ\Phi is a morphism of Hodge structures of bidegree (−1,−1)(-1,-1) because it is induced by an algebraic correspondence. Finally, if v∈TX,Qv\in T_{X,\mathbb{Q}} and D∈NS⁡(F)⊗QD\in\operatorname{NS}(F)\otimes\mathbb{Q}, then (Φv,D)F=(v,e∗p∗D)X=0(\Phi v,D)_F=(v,e_*p^*D)_X=0, because e∗p∗De_*p^*D is an algebraic class: so Φ\Phi maps TX,QT_{X,\mathbb{Q}} to TF,QT_{F,\mathbb{Q}}.

The map Φ\Phi is the transpose of the cylinder map e∗p∗:H2(F,Q)→H4(X,Q)e_*p^*:H^2(F,\mathbb{Q})\to H^4(X,\mathbb{Q}) studied by Lewis in [33, 34] for hypersurfaces of degree nn in Pn\mathbb{P}^n; see also [38]. For these hypersurfaces, where aY=n!a_Y=n!, the eigenvalue λ=−n!q\lambda=-n!q is due to Collino–Jinzenji [9 Proposition 2].

Quantum Cohomology of Küchle fourfolds

In this section, we compute the small quantum Euler operator UX(0)=H⋆U_X(0)=H\star of the Küchle fourfolds of types (c1), (c6), (b8), and (b11). In each case, Quantum Lefschetz gives the restriction of UX(0)U_X(0) to the inherited cohomology BB, and the results of Section 2 then determine its action on the vanishing cohomology VV. We briefly recall Quantum Lefschetz in §4.1 and outline the method in §4.2.

Quantum Lefschetz

We use Quantum Lefschetz in the form of [10 Theorem 1.1], which relates Givental’s Lagrangian cone of a zero locus to the Euler-twisted cone of the ambient variety. In this subsection only, the inclusion i:X↪Yi:X\hookrightarrow Y denotes the zero locus of a regular section of a convex vector bundle E\mathcal{E} on a smooth projective variety YY.

The genus-zero Gromov–Witten invariants of XX are encoded by a Lagrangian cone LX\mathcal{L}_X in the symplectic space

HX=H∙(X;ΛXNov)⊗C((z−1)),ΩX(f,g)=Res⁡z=0(f(−z),g(z)) dz.\mathcal{H}_X=H^\bullet(X;\Lambda_X^{\mathrm{Nov}})\otimes\mathbb{C}((z^{-1})),\qquad\Omega_X(f,g)=\operatorname{Res}_{z=0}(f(-z),g(z))\,\mathrm{d}z.

Here ΛXNov\Lambda_X^{\mathrm{Nov}} is the Novikov ring whose monomials QβQ^\beta record degrees of effective curve classes β∈NE(X)⊂H2(X,Z)\beta\in\mathrm{NE}(X)\subset H_2(X,\mathbb{Z}).

A general point of LX⊂HX\mathcal{L}_X\subset\mathcal{H}_X has the form

JX(t):=−z1X+t(z)+∑β,n,m,k,ϵQβn!⟨tk1ψ1k1,…,tknψnkn,ϕϵψn+1m⟩0,n+1,βXϕϵ(−z)−m−1,(9)J_X(\mathbf{t}):=-z\mathbf{1}_X+\mathbf{t}(z)+\sum_{\beta,n,m,k,\epsilon}\frac{Q^\beta}{n!}\left\langle t_{k_1}\psi_1^{k_1},\ldots,t_{k_n}\psi_n^{k_n},\phi_\epsilon\psi_{n+1}^{m}\right\rangle^{X}_{0,n+1,\beta}\phi^\epsilon(-z)^{-m-1}, \tag*{(9)}

where t(z)=t0+t1z+t2z2+⋯\mathbf{t}(z)=t_0+t_1z+t_2z^2+\cdots with tk∈H∙(X;ΛXNov)t_k\in H^\bullet(X;\Lambda_X^{\mathrm{Nov}}), and {ϕϵ}\{\phi_\epsilon\} and {ϕϵ}\{\phi^\epsilon\} are Poincaré dual bases of H∙(X;Q)H^\bullet(X;\mathbb{Q}). The JJ-function JX(τ,z)J_X(\tau,z) is the unique point of LX\mathcal{L}_X of the form −z+τ+O(z−1)-z+\tau+O(z^{-1}).

Replacing the invariants of XX by the invariants of YY twisted by the S1S^1-equivariant Euler class e\mathbf{e} of E\mathcal{E}, with equivariant parameter κ\kappa, gives a cone LY,e\mathcal{L}_{Y,\mathbf{e}} with a general point JY,eJ_{Y,\mathbf{e}}. It lives in the corresponding equivariant symplectic space for YY with the twisted pairing

(a,b)Y,e=∫Ya⌣b⌣e(E).(a,b)_{Y,\mathbf{e}}=\int_Y a\smile b\smile\mathbf{e}(\mathcal{E}).

Theorem 4.1 (Quantum Lefschetz, [10 Theorem 1.1]). Let i:X↪Yi:X\hookrightarrow Y be the smooth zero locus of a regular section of a convex vector bundle E→Y\mathcal{E}\to Y. The non-equivariant limit of JY,e(t)J_{Y,\mathbf{e}}(\mathbf{t}) exists and satisfies

i∗JY,e(t)∣κ=0=JX(i∗t).(10)i^*J_{Y,\mathbf{e}}(\mathbf{t})\big|_{\kappa=0}=J_X(i^*\mathbf{t}). \tag*{(10)}

In particular, i∗LY,e∣κ=0⊆LXi^*\mathcal{L}_{Y,\mathbf{e}}\big|_{\kappa=0}\subseteq\mathcal{L}_X. Throughout the homomorphism Qδ↦Qi∗δQ^\delta\mapsto Q^{i_*\delta} is applied to ΛXNov\Lambda_X^{\mathrm{Nov}}.

Let {ϕα}\{\phi_\alpha\} be a homogeneous basis of H∙(X,Q)H^\bullet(X,\mathbb{Q}), with ϕα∈H2dα(X,Q)\phi_\alpha\in H^{2d_\alpha}(X,\mathbb{Q}), and write τ=∑αταϕα\tau=\sum_\alpha\tau^\alpha\phi_\alpha. We recover the big quantum product ⋆τ\star_\tau from the JJ-function by Coates–Givental [7 Equation 16]:

−z∂2JX∂τα∂τβ=∑γAα,βγ(τ)∂JX∂τγ,ϕα⋆τϕβ=∑γAα,βγ(τ)ϕγ.(11)-z\frac{\partial^2J_X}{\partial\tau^\alpha\partial\tau^\beta}=\sum_\gamma A^\gamma_{\alpha,\beta}(\tau)\frac{\partial J_X}{\partial\tau^\gamma},\qquad\phi_\alpha\star_\tau\phi_\beta=\sum_\gamma A^\gamma_{\alpha,\beta}(\tau)\phi_\gamma. \tag*{(11)}

This formulation is particularly useful when XX inherits classes from YY that are not powers of the hyperplane class, as in our examples. It provides the input we need to compute the Hodge atom birational invariants of Katzarkov–Kontsevich–Pantev–Yu in [27].

As discussed in the introduction, to use these invariants we study the generalized eigenspaces of quantum multiplication by the Euler vector field

Eu⁡X(τ)=c1(X)+∑α(1−dα)ταϕα,UX(τ)(a)=Eu⁡X(τ)⋆τa.\operatorname{Eu}_{X}(\tau)=c_{1}(X)+\sum_{\alpha}(1-d_{\alpha})\tau^{\alpha}\phi_{\alpha},\qquad U_{X}(\tau)(a)=\operatorname{Eu}_{X}(\tau)\star_{\tau}a.

At τ=0\tau=0, the product ⋆0\star_{0} is the small quantum product of § 2. For the index-one, Picard-rank one fourfolds considered in this paper, we write Qβ=qH⋅βQ^{\beta}=q^{H\cdot\beta} and obtain

UX(0)=c1(X)⋆=H⋆.(12)U_{X}(0)=c_{1}(X)\star=H\star. \tag*{(12)}

Method

Let G=Gr⁡(k,n)G=\operatorname{Gr}(k,n) be the Grassmannian of kk-dimensional subspaces of Cn\mathbb{C}^{n}, with tautological subbundle SS, quotient bundle QQ, and Plücker line bundle OG(1)=det⁡(S∨)\mathcal{O}_{G}(1)=\det(S^{\vee}). We write ZM(F)Z_{M}(\mathcal{F}) for the zero locus of a general regular section of the vector bundle F\mathcal{F} on a variety MM. Each fourfold considered below has the form

X=ZY(E),Y=ZG(F),(13)X=Z_{Y}(\mathcal{E}),\qquad Y=Z_{G}(\mathcal{F}), \tag*{(13)}

where F\mathcal{F} is a globally generated homogeneous vector bundle on GG and E\mathcal{E} is a direct sum of copies of OY(1)\mathcal{O}_{Y}(1); see Table 2.

Küchle TypeGGF\mathcal{F}E\mathcal{E}h3,1(X)h^{3,1}(X)h2,2(X)h^{2,2}(X)b4(X)b_{4}(X)
(b8)Gr⁡(2,7)\operatorname{Gr}(2,7)Sym⁡2S∨\operatorname{Sym}^{2}S^{\vee}OY(1)⊕3\mathcal{O}_{Y}(1)^{\oplus3}33844
(b11)Gr⁡(2,8)\operatorname{Gr}(2,8)Q∨(1)Q^{\vee}(1)OY(1)⊕2\mathcal{O}_{Y}(1)^{\oplus2}23135
(c1)Gr⁡(3,6)\operatorname{Gr}(3,6)—OY(1)⊕5\mathcal{O}_{Y}(1)^{\oplus5}55262
(c6)Gr⁡(3,7)\operatorname{Gr}(3,7)(∧2S∨)⊕2(\wedge^{2}S^{\vee})^{\oplus2}OY(1)⊕2\mathcal{O}_{Y}(1)^{\oplus2}23034

Table 2. Küchle fourfolds from [Küc95, Theorem 3] and their middle Hodge numbers.

By Küchle [Küc95, Theorem 3], XX is a Fano fourfold of index one, so that

−KX=H=c1(OG(1))∣X=c1(X).-K_{X}=H=c_{1}(\mathcal{O}_{G}(1))|_{X}=c_{1}(X).

It satisfies the standing assumptions of § 2, namely, it is Picard-rank one with Hodd(X,Q)=0H^{\mathrm{odd}}(X,\mathbb{Q})=0.

Now let Y′Y' be a general five-dimensional linear section of YY containing XX. It is a smooth fivefold with hyperplane section XX, and we apply § 2 to X⊂Y′X\subset Y'. By the Lefschetz hyperplane theorem Hj(Y,Q)→Hj(Y′,Q)H^{j}(Y,\mathbb{Q})\to H^{j}(Y',\mathbb{Q}) is an isomorphism for j≤4j\leq4, while H6(X,Q)H^{6}(X,\mathbb{Q}) and H8(X,Q)H^{8}(X,\mathbb{Q}) are spanned by H3H^{3} and H4H^{4}. Hence the inherited cohomology BXB_{X} of § 2 is the image of H∙(Y,Q)H^{\bullet}(Y,\mathbb{Q}). We put BX4:=BX∩H4(X,Q)B_{X}^{4}:=B_{X}\cap H^{4}(X,\mathbb{Q}), so that VXV_{X} is the orthogonal complement of BX4B_{X}^{4} in H4(X,Q)H^{4}(X,\mathbb{Q}).

Lemma 4.2. Suppose that BX4B_{X}^{4} is spanned by algebraic classes. Then

h3,1(VX)=h3,1(X),h2,2(VX)=h2,2(X)−dim⁡BX4,h^{3,1}(V_{X})=h^{3,1}(X),\qquad h^{2,2}(V_{X})=h^{2,2}(X)-\dim B_{X}^{4},

and the Poincaré pairing on VXV_{X} has signature (h2,2(VX),2h3,1(VX))(h^{2,2}(V_{X}),2h^{3,1}(V_{X})).

Proof. Since XX is Fano, h4,0(X)=0h^{4,0}(X)=0. The subspace VXV_{X} is the orthogonal complement of BX4B_{X}^{4}, which is of type (2,2)(2,2); hence it is a sub-Hodge structure of H4(X,Q)H^{4}(X,\mathbb{Q}) with the stated Hodge numbers. Let v∈VXv\in V_{X}. Since H2∈BX4H^{2}\in B_{X}^{4} we have (H⌣v,H)=(v,H2)=0(H\smile v,H)=(v,H^{2})=0, and therefore H⌣v=0H\smile v=0. Thus VXV_{X} consists of primitive classes, and Hodge–Riemann gives the stated signature relations. □

Corollary 4.3. Suppose that the middle inherited cohomology BX4B_X^4 is spanned by algebraic classes and that h3,1(VX)>0h^{3,1}(V_X)>0 and h2,2(VX)>0h^{2,2}(V_X)>0. Then BB is stable under H⋆H\star, and there is λ∈K\lambda\in K with the following properties.

  1. H⋆v=λvH\star v=\lambda v for all v∈Vv\in V, and λ\lambda is a repeated root of det⁡(t−UX(0)∣B)\det(t-U_X(0)|_B).

  2. There is a unique class bχ∈Bb_\chi\in B such that v⋆w=(v,w)bχv\star w=(v,w)b_\chi for all v,w∈Vv,w\in V. It is isotropic and satisfies H⋆bχ=λbχH\star b_\chi=\lambda b_\chi and (bχ,1)=1(b_\chi,1)=1.

  3. det⁡(t−UX(0))=(t−λ)dim⁡VXdet⁡(t−UX(0)∣B)\det(t-U_X(0))=(t-\lambda)^{\dim V_X}\det(t-U_X(0)|_B).

Proof. By Lemma 4.2, the Poincaré pairing on VXV_X is indefinite. Hence Proposition 2.1 applies to X⊂Y′X\subset Y', and the small quantum product is O(V)O(V)-equivariant. Proposition 2.2 with λ=χ(H)\lambda=\chi(H) shows that BB is invariant under H⋆H\star, and also gives (1) and (2): because the pairing on VV is nonzero, the class bχb_\chi is unique. Part (3) follows from (1). □

Inherited matrix computation. Assume that Corollary 4.3 applies, and let BX=(φ1,…,φs)\mathcal{B}_X=(\varphi_1,\ldots,\varphi_s) be a homogeneous basis of BX\mathcal{B}_X, with φj∈H2dj(X,Q)\varphi_j\in H^{2d_j}(X,\mathbb{Q}).

As XX has index one, qq has degree one, and the degree axiom gives

H⋆φj=∑i=1saijq1+dj−diφi,aij∈Q,(14)H\star\varphi_j=\sum_{i=1}^{s}a_{ij}q^{1+d_j-d_i}\varphi_i,\qquad a_{ij}\in\mathbb{Q}, \tag*{(14)}

with aij=0a_{ij}=0 whenever di>dj+1d_i>d_j+1. We call

AX(q):=(aijq1+dj−di)i,jA_X(q):=\left(a_{ij}q^{1+d_j-d_i}\right)_{i,j}

the inherited matrix of XX in the basis BX\mathcal{B}_X; it is the matrix of UX(0)∣BU_X(0)|_B.

Writing D=diag⁡(qd1,…,qds)D=\operatorname{diag}(q^{d_1},\ldots,q^{d_s}), we have AX(q)=qD−1AX(1)DA_X(q)=qD^{-1}A_X(1)D, so that

det⁡(t−UX(0)∣B)=qsdet⁡((t/q)Is−AX(1)).(15)\det\left(t-U_X(0)|_B\right)=q^s\det\left((t/q)I_s-A_X(1)\right). \tag*{(15)}

Note that the entries of this matrix, for di=dj+1d_i=d_j+1, are the structure constants of the cup product with HH. Moreover, H⋆H\star is self-adjoint for the Poincaré pairing, so GXAX(q)G_XA_X(q) is symmetric, where GXG_X is the Gram matrix of BX\mathcal{B}_X.

To compute AX(q)A_X(q), we apply Theorem 4.1 to a family IY,eI_{Y,e} of points of the twisted cone from § 4.1. This gives

IX(q,z)=i∗IY,e(q,z;κ)∣κ=0∈LX⊂HX,I_X(q,z)=\left.i^*I_{Y,e}(q,z;\kappa)\right|_{\kappa=0}\in\mathcal{L}_X\subset\mathcal{H}_X,

and in each case IX(q,z)=−z+aXq+O(z−1)I_X(q,z)=-z+a_Xq+O(z^{-1}) for some aX∈Qa_X\in\mathbb{Q}. By the string equation and the characterization of JXJ_X in §4.1,

IX(q,z)=JX(aXq1;q,z)=e−aXq/zJX(0;q,z).(16)I_X(q,z)=J_X(a_Xq1;q,z)=e^{-a_Xq/z}J_X(0;q,z). \tag*{(16)}

Writing τ=∑itiφi\tau=\sum_i t^i\varphi_i and applying (11) to our normalization of the family JX(τ;q,z)J_X(\tau;q,z) at τ=0\tau=0 gives AX(q)A_X(q). Equivalently, if A~X(q)\widetilde{A}_X(q) is the matrix of UX(aXq1)∣BU_X(a_Xq1)|_B, then

AX(q)=A~X(q)−aXqIs.(17)A_X(q)=\widetilde{A}_X(q)-a_XqI_s. \tag*{(17)}

Remark 4.4 (String shift and the vanishing eigenvalue). Let aXa_X be the coefficient of the unit-coordinate correction in (4.8). We show that it is the number of lines of Y′Y' through a general point. Indeed, the hypergeometric modification for the last hyperplane section multiplies the degree-dd term of JY′(0;q,z)J_{Y'}(0;q,z) by ∏m=1d(H−mz)\prod_{m=1}^{d}(H-mz). Since Y′Y' has index two, the coefficient of qq in JY′(0;q,z)J_{Y'}(0;q,z) is −az−11+O(z−2)-az^{-1}1+O(z^{-2}) with a=⟨[pt],1⟩0,1,1Y′a=\langle[\mathrm{pt}],1\rangle^{Y'}_{0,1,1}, and multiplying by H−zH-z gives

aX=⟨[pt],1⟩0,1,1Y′.(18)a_X=\langle[\mathrm{pt}],1\rangle^{Y'}_{0,1,1}. \tag*{(18)}

Thus the unit-coordinate correction is aXq1a_Xq\mathbf{1}, which agrees with Iritani’s mirror-map formula for a hyperplane section of an index-two Fano variety [24 Section 0.2].

Because (18) is deformation invariant, this holds for every smooth pair X⊂Y′X \subset Y' of the given type. For types (b8), (b11), (c1), (c6) we have aX=4,3,6,5a_X=4,3,6,5, which agree with the vanishing eigenvalues λ=−aXq\lambda=-a_Xq computed in §4.

Consequently, the translation in (17) gives an operator annihilating the vanishing summand:

UX(aXq1)∣VX⊗K=0so thatdet⁡(tI−UX(aXq1))=tdim⁡VXdet⁡(tIs−A~X(q)).U_X(a_Xq\mathbf{1})|_{V_X\otimes K}=0 \qquad\text{so that} \qquad\det(tI-U_X(a_Xq\mathbf{1}))=t^{\dim V_X}\det(tI_s-\widetilde A_X(q)).

This gives a cohomological interpretation of what Golyshev refers to as the Givental constant in [17 Section 1.2]. There, he explains that the Quantum Lefschetz comparison of spectra requires an additional translation when the hyperplane section has index one. After this translation, the vanishing cohomology contributes only zero eigenvalues to the shifted operator, which is the behavior suggested by Iritani’s question on extending the Quantum Lefschetz comparison of characteristic polynomials to include primitive cohomology [24 Question 15].

Applications to Küchle fourfolds

Küchle fourfold of type (b8)

Let G=Gr⁡(2,7)G=\operatorname{Gr}(2,7) and

Y=ZG(Sym⁡2S∨),E=OY(1)⊕3,X=ZY(E).Y=Z_G(\operatorname{Sym}^2S^\vee), \qquad E=\mathcal{O}_Y(1)^{\oplus3}, \qquad X=Z_Y(E).

Put P=c2(S∨)∣XP=c_2(S^\vee)|_X. The inherited cohomology BXB_X has the basis

Bb8=(1,H,H2,P,H3,H4),B_{b8}=(1,H,H^2,P,H^3,H^4),

and its Poincaré pairing is determined by the intersection numbers

H4=56,H2⋅P=20,P2=8.H^4=56, \qquad H^2\cdot P=20, \qquad P^2=8.

Since the classes H2H^2 and PP are algebraic, Lemma 4.2 gives h3,1(VX)=3h^{3,1}(V_X)=3 and h2,2(VX)=36h^{2,2}(V_X)=36. In particular, dim⁡VX=42\dim V_X=42 and Corollary 4.3 applies.

Proposition 4.5 (Type (b8)). Let XX be a Küchle fourfold of type (b8).

  1. In the basis Bb8B_{b8}, the matrix of UX(0)∣BU_X(0)|_B is

Ab8(q)=(096q22592q3936q351968q4434560q516q300q2108q26048q351968q4016q4q280q22352q3008q−2q56q2672q30015146q96q2000010).A_{b8}(q)= \begin{pmatrix} 0 & 96q^2 & 2592q^3 & 936q^3 & 51968q^4 & 434560q^5 \\ 1 & 6q & 300q^2 & 108q^2 & 6048q^3 & 51968q^4 \\ 0 & 1 & 6q & 4q & 280q^2 & 2352q^3 \\ 0 & 0 & 8q & -2q & 56q^2 & 672q^3 \\ 0 & 0 & 1 & \frac{5}{14} & 6q & 96q^2 \\ 0 & 0 & 0 & 0 & 1 & 0 \end{pmatrix}.

Consequently,

det⁡(t−UX(0)∣B)=(t+4q)3(t+8q)(t2−36qt−176q2).\det(t-U_X(0)|_B)=(t+4q)^3(t+8q)(t^2-36qt-176q^2).
  1. For all v,w∈Vv,w\in V,

H⋆v=−4qv,andv⋆w=(v,w)bχ,H\star v=-4qv, \qquad\text{and} \qquad v\star w=(v,w)_{b_\chi},

where bχ=b/56b_\chi=b/56 and the class

b=H4−4qH3−56q2H2−672q3H−4256q4b=H^4-4qH^3-56q^2H^2-672q^3H-4256q^4

spans the (−4q)(-4q)-eigenspace of UX(0)∣BU_X(0)|_B.

(3) We have

det⁡(t−UX(0))=(t+4q)45(t+8q)(t2−36qt−176q2).\det(t-U_X(0))=(t+4q)^{45}(t+8q)(t^2-36qt-176q^2).

In particular MX=45M_X=45.

Proof. (1) Apply the method of § 4.2 to the Euler-twisted Grassmannian II-function of GG, twisted by Sym⁡2S∨⊕OG(1)⊕3\operatorname{Sym}^2 S^\vee\oplus\mathcal{O}_G(1)^{\oplus3}, constructed using the abelian/non-abelian correspondence of Inoue–Ito–Miura [22 Definition 3.3, Theorem 3.4]. This gives Ab8(q)A_{b8}(q) and the characteristic polynomial follows.

(2) By Proposition 2.2, H⋆H\star acts on VV by a scalar which is a repeated root of the polynomial from the previous step. Hence λ=−4q\lambda=-4q. A direct computation with Ab8(q)A_{b8}(q) shows that the (−4q)(-4q)-eigenspace of UX(0)∣BU_X(0)|_B is the line spanned by bb and (b,1)=H4=56(b,1)=H^4=56. By Proposition 2.2, this line contains bχb_\chi, and (bχ,1)=1(b_\chi,1)=1. Hence bχ=b/56b_\chi=b/56.

(3) This is Corollary 4.3 since dim⁡VX=42\dim V_X=42. □

Küchle fourfold of type (b11)

Let G=Gr⁡(2,8)G=\operatorname{Gr}(2,8) and

Y=ZG(Q(1)),E=OY(1)⊕2,X=ZY(E).Y=Z_G\bigl(Q^{(1)}\bigr),\qquad\mathcal{E}=\mathcal{O}_Y(1)^{\oplus2},\qquad X=Z_Y(\mathcal{E}).

Let U1,U2\mathcal{U}_1,\mathcal{U}_2 be Meng’s universal bundles on YY, of ranks two and three [36], and put

c2=c2(U2∨),d2=c2(U1∨),HY=c1(U1∨)=c1(U2∨).c_2=c_2(\mathcal{U}_2^\vee),\qquad d_2=c_2(\mathcal{U}_1^\vee),\qquad H_Y=c_1(\mathcal{U}_1^\vee)=c_1(\mathcal{U}_2^\vee).

Define

e1=−c2+2d2,e2=HY2−2d2,e3=c2−d2,eˉj=ej∣X.e_1=-c_2+2d_2,\qquad e_2=H_Y^2-2d_2,\qquad e_3=c_2-d_2,\qquad\bar e_j=e_j|_X.

The inherited cohomology BXB_X has the basis

Bb11=(1,H,eˉ1,eˉ2,eˉ3,H3,pt),pt:=H457\mathcal{B}_{b11}=(1,H,\bar e_1,\bar e_2,\bar e_3,H^3,\mathrm{pt}),\qquad\mathrm{pt}:=\frac{H^4}{57}

and its Poincaré pairing is determined by H4=57H^4=57, the relation H2=2eˉ1+eˉ2+2eˉ3H^2=2\bar e_1+\bar e_2+2\bar e_3 and the Gram matrix of (eˉ1,eˉ2,eˉ3)(\bar e_1,\bar e_2,\bar e_3), which is

(231393132).\begin{pmatrix} 2 & 3 & 1\\ 3 & 9 & 3\\ 1 & 3 & 2 \end{pmatrix}.

Since the classes eˉj\bar e_j are algebraic, Lemma 4.2 gives h3,1(VX)=2h^{3,1}(V_X)=2 and h2,2(VX)=28h^{2,2}(V_X)=28. In particular, dim⁡VX=32\dim V_X=32 and Corollary 4.3 applies.

Proposition 4.6 (Type (b11)). Let XX be a Küchle fourfold of type (b11).

(1) In the basis Bb11\mathcal{B}_{b11}, the matrix of UX(0)∣BU_X(0)|_B is

Ab11(q)=(084q2333q3846q3333q342408q45805q515q42q2108q242q25301q3744q402−2q3q2q342q251q3013q6q3q456q260q3022q3q−2q342q251q3003/197/193/195q28q2/1900000570).A_{b11}(q)= \begin{pmatrix} 0 & 84q^2 & 333q^3 & 846q^3 & 333q^3 & 42408q^4 & 5805q^5\\ 1 & 5q & 42q^2 & 108q^2 & 42q^2 & 5301q^3 & 744q^4\\ 0 & 2 & -2q & 3q & 2q & 342q^2 & 51q^3\\ 0 & 1 & 3q & 6q & 3q & 456q^2 & 60q^3\\ 0 & 2 & 2q & 3q & -2q & 342q^2 & 51q^3\\ 0 & 0 & 3/19 & 7/19 & 3/19 & 5q & 28q^2/19\\ 0 & 0 & 0 & 0 & 0 & 57 & 0 \end{pmatrix}.

Consequently,

det⁡(t−UX(0)∣B)=(t+3q)3(t+4q)2(t2−29qt−339q2).\det(t-U_X(0)|_B)=(t+3q)^3(t+4q)^2(t^2-29qt-339q^2).

(2) For all v,w∈Vv,w \in V,

H⋆v=−3qv,andv⋆w=(v,w)bχ,H \star v = -3qv, \qquad\text{and} \qquad v \star w = (v,w)b_{\chi},

where bχ=b/57b_{\chi}=b/57 and the class

b=H4−3qH3−57q2(eˉ1+2eˉ2+eˉ3)−684q3H−3933q4b=H^{4}-3qH^{3}-57q^{2}(\bar e_{1}+2\bar e_{2}+\bar e_{3})-684q^{3}H-3933q^{4}

spans the (−3q)(-3q)-eigenspace of UX(0)∣BU_{X}(0)|_{B}.

(3) We have

det⁡(t−UX(0))=(t+3q)35(t+4q)2(t2−29qt−339q2).\det(t-U_{X}(0))=(t+3q)^{35}(t+4q)^{2}(t^{2}-29qt-339q^{2}).

In particular MX=35M_{X}=35.

Proof. (1) We obtain JYJ_{Y} from Meng’s quantum Chevalley formula [36 Corollary 5.26], and apply Coates–Givental hypergeometric modification [7 Theorem 2] for OY(1)⊕2\mathcal{O}_{Y}(1)^{\oplus2}. The method of §4.2 then gives Ab11(q)A_{b11}(q) and the characteristic polynomial follows.

(2) By Corollary 4.3, H⋆H\star acts on VV by a scalar λ\lambda, which is a repeated root of the polynomial in (1). The repeated roots are −3q-3q and −4q-4q: the factor t2−29qt−339q2t^{2}-29qt-339q^{2} has discriminant 2197q22197q^{2}, and does not vanish at t=−3qt=-3q nor at t=−4qt=-4q.

We exclude −4q-4q because a computation with Ab11(q)A_{b11}(q) shows that the (−4q)(-4q)-eigenspace of UX(0)∣BU_{X}(0)|_{B} is the plane spanned by

ξ=eˉ1−eˉ3andη=2H4−8qH3−57q2(3eˉ1+2eˉ2+3eˉ3)−1026q3H−6498q4,\xi=\bar e_{1}-\bar e_{3} \qquad\text{and}\qquad \eta=2H^{4}-8qH^{3}-57q^{2}(3\bar e_{1}+2\bar e_{2}+3\bar e_{3})-1026q^{3}H-6498q^{4},

and that

(ξ,ξ)=2,(ξ,η)=0,(η,η)=−6(57q2)2.(\xi,\xi)=2,\qquad(\xi,\eta)=0,\qquad(\eta,\eta)=-6(57q^{2})^{2}.

Hence, for α,β∈K\alpha,\beta\in K,

(αξ+βη,αξ+βη)=2(α2−3(57q2β)2),(\alpha\xi+\beta\eta,\alpha\xi+\beta\eta)=2\left(\alpha^{2}-3(57q^{2}\beta)^{2}\right),

which vanishes only for α=β=0\alpha=\beta=0 (because 33 is not a square in KK). Therefore the eigenspace contains no nonzero isotropic vector, whereas bχb_{\chi} is isotropic. Hence λ=−3q\lambda=-3q.

A computation with Ab11(q)A_{b11}(q) shows that the (−3q)(-3q)-eigenspace of UX(0)∣BU_{X}(0)|_{B} is the line spanned by bb, and (b,1)=H4=57(b,1)=H^{4}=57. As this line contains bχb_{\chi} and (bχ,1)=1(b_{\chi},1)=1, we get bχ=b/57b_{\chi}=b/57.

(3) This is Corollary 4.3 since dim⁡VX=32\dim V_{X}=32. □

Küchle fourfold of type (c1)

Let Y=G=Gr⁡(3,6)Y=G=\operatorname{Gr}(3,6) and

E=OY(1)⊕5,X=ZY(E).\mathcal{E}=\mathcal{O}_{Y}(1)^{\oplus5}, \qquad X=Z_{Y}(\mathcal{E}).

Put P=c2(S∨)∣XP=c_{2}(S^{\vee})|_{X}. The inherited cohomology BXB_{X} has the basis

Bc1=(1,H,H2,P,H3,H4),\mathcal{B}_{\mathrm{c1}}=(1,H,H^{2},P,H^{3},H^{4}),

and its Poincaré pairing is determined by the intersection numbers

H4=42,H2⋅P=21,P2=11.H^{4}=42,\qquad H^{2}\cdot P=21,\qquad P^{2}=11.

Since H2H^{2} and PP are algebraic, Lemma 4.2 gives h3,1(VX)=5h^{3,1}(V_{X})=5 and h2,2(VX)=50h^{2,2}(V_{X})=50. In particular dim⁡VX=60\dim V_{X}=60, and Corollary 4.3 applies.

Proposition 4.7 (Type (c1)). Let XX be a Küchle fourfold of type (c1).

(1) In the basis Bc1\mathcal{B}_{c_1}, the matrix of UX(0)∣BU_X(0)|_B is

Ac1(q)=(0180q26804q33402q3184464q42026080q5110q588q2294q216380q3184464q40115q10q588q26804q3000−5q000011210q180q2000010).A_{c_1}(q)= \begin{pmatrix} 0 & 180q^2 & 6804q^3 & 3402q^3 & 184464q^4 & 2026080q^5 \\ 1 & 10q & 588q^2 & 294q^2 & 16380q^3 & 184464q^4 \\ 0 & 1 & 15q & 10q & 588q^2 & 6804q^3 \\ 0 & 0 & 0 & -5q & 0 & 0 \\ 0 & 0 & 1 & \frac{1}{2} & 10q & 180q^2 \\ 0 & 0 & 0 & 0 & 1 & 0 \end{pmatrix}.

Consequently

det⁡(t−UX(0)∣B)=(t+6q)3(t+5q)2(t−58q).\det(t-U_X(0)|_B)=(t+6q)^3(t+5q)^2(t-58q).

(2) For all v,w∈Vv,w\in V,

H⋆v=−6qvandv⋆w=(v,w)bX,H\star v=-6qv \qquad\text{and}\qquad v\star w=(v,w)b_X,

where bX=b/42b_X=b/42 and the class

b=H4−6qH3−84q2H2−1512q3H−12600q4b=H^4-6qH^3-84q^2H^2-1512q^3H-12600q^4

spans the (−6q)(-6q)-eigenspace of UX(0)∣BU_X(0)|_B.

(3) We have

det⁡(t−UX(0))=(t+6q)63(t+5q)2(t−58q).\det(t-U_X(0))=(t+6q)^{63}(t+5q)^2(t-58q).

In particular MX=63M_X=63.

Proof. (1) Apply the method of § 4.2 to the Euler-twisted II-function of GG, twisted by OG(1)⊕5\mathcal{O}_G(1)^{\oplus5}, using the abelian/non-abelian correspondence of Inoue–Ito–Miura [22 Definition 3.3, Theorem 3.4]. This gives Ac1(q)A_{c_1}(q), and the characteristic polynomial follows.

(2) By Corollary 4.3, H⋆H\star acts on VV by a scalar λ\lambda which is a repeated root of the polynomial in (1), so λ∈{−6q,−5q}\lambda\in\{-6q,-5q\}. We exclude −5q-5q as in the proof of Proposition 4.6: a computation with Ac1(q)A_{c_1}(q) shows that the (−5q)(-5q)-eigenspace of UX(0)∣BU_X(0)|_B is the plane spanned by

ζ=H2−2Pandζ′=H4−5qH3−105q2H2−1764q3H−14364q4,\zeta=H^2-2P \qquad\text{and}\qquad\zeta'=H^4-5qH^3-105q^2H^2-1764q^3H-14364q^4,

and that (ζ,ζ)=2(\zeta,\zeta)=2, (ζ,ζ′)=0(\zeta,\zeta')=0 and (ζ′,ζ′)=−6(21q)2(\zeta',\zeta')=-6(21q)^2. Since 3 is not a square in KK, this plane contains no nonzero isotropic vector, whereas bXb_X is isotropic. Hence λ=−6q\lambda=-6q, in agreement with Remark 4.4, where aX=6a_X=6. The (−6q)(-6q)-eigenspace of UX(0)∣BU_X(0)|_B is the line spanned by bb and (b,1)=H4=42(b,1)=H^4=42. As this line contains bXb_X and (bX,1)=1(b_X,1)=1, we get bX=b/42b_X=b/42.

(3) This is Corollary 4.3 since dim⁡VX=60\dim V_X=60.

Küchle fourfold of type (c6)

Let G=Gr⁡(3,7)G=\operatorname{Gr}(3,7) and

Y=ZG((⋀2S∨)⊕2),E=OY(1)⊕2,X=ZY(E).Y=Z_G\left(\left(\bigwedge^2 S^\vee\right)^{\oplus2}\right), \qquad\mathcal{E}=\mathcal{O}_Y(1)^{\oplus2}, \qquad X=Z_Y(\mathcal{E}).

Put P=c2(Q)∣XP=c_2(Q)|_X. The inherited cohomology BXB_X has the basis

Bc6=(1,H,H2,P,H3,H4),\mathcal{B}_{c_6}=(1,H,H^2,P,H^3,H^4),

and its Poincaré pairing is determined by the intersection numbers

H4=61,H2⋅P=33,P2=18.H^4=61,\qquad H^2\cdot P=33,\qquad P^2=18.

Since H2H^2 and PP are algebraic, Lemma 4.2 gives h3,1(VX)=2h^{3,1}(V_X)=2 and h2,2(VX)=28h^{2,2}(V_X)=28. In particular dim⁡VX=32\dim V_X=32, and Corollary 4.3 applies.

Proposition 4.8 (Type (c6)). Let XX be a Küchle fourfold of type (c6).

(1) In the basis Bc6\mathcal{B}_{c6}, the matrix of UX(0)∣BU_X(0)|_B is

Ac6(q)=(088q22097q31134q338308q4311710q516q244q2132q24575q338308q40113q9q244q22196q300−8q−8q0−183q300133616q88q2000010).A_{c6}(q)= \begin{pmatrix} 0 & 88q^2 & 2097q^3 & 1134q^3 & 38308q^4 & 311710q^5 \\ 1 & 6q & 244q^2 & 132q^2 & 4575q^3 & 38308q^4 \\ 0 & 1 & 13q & 9q & 244q^2 & 2196q^3 \\ 0 & 0 & -8q & -8q & 0 & -183q^3 \\ 0 & 0 & 1 & \frac{33}{61} & 6q & 88q^2 \\ 0 & 0 & 0 & 0 & 1 & 0 \end{pmatrix}.

Consequently

det⁡(t−UX(0)∣B)=(t+5q)3(t3−32qt2−195q2t−293q3).\det(t-U_X(0)|_B)=(t+5q)^3(t^3-32qt^2-195q^2t-293q^3).

(2) For all v,w∈Vv,w\in V,

H⋆v=−5q vandv⋆w=(v,w)bχ,H\star v=-5q\,v \qquad\text{and}\qquad v\star w=(v,w)b_\chi,

where bχ=b/61b_\chi=b/61 and the class

b=H4−5qH3−61q2P−427q3H−2684q4b=H^4-5qH^3-61q^2P-427q^3H-2684q^4

spans the (−5q)(-5q)-eigenspace of UX(0)∣BU_X(0)|_B.

(3) We have

det⁡(t−UX(0))=(t+5q)35(t3−32qt2−195q2t−293q3).\det(t-U_X(0))=(t+5q)^{35}(t^3-32qt^2-195q^2t-293q^3).

In particular MX=35M_X=35.

Proof. (1) Apply the method of § 4.2 to the Euler-twisted Grassmannian II-function of GG, twisted by (⋀2S∨)⊕2⊕OG(1)⊕2\left(\bigwedge^2 S^\vee\right)^{\oplus2}\oplus\mathcal{O}_G(1)^{\oplus2}, using the abelian/non-abelian correspondence of Inoue–Ito–Miura [22 Definition 3.3, Theorem 3.4]. This gives Ac6(q)A_{c6}(q), and the characteristic polynomial follows.

(2) By Corollary 4.3, H⋆H\star acts on VV by a scalar λ\lambda, which is a repeated root of the polynomial in (1). Its cubic factor is squarefree over Q(q)\mathbb{Q}(q), and gives the nonzero value −243q3-243q^3 at t=−5qt=-5q. Thus −5q-5q is the unique repeated root and λ=−5q\lambda=-5q.

A computation with Ac6(q)A_{c6}(q) shows that the (−5q)(-5q)-eigenspace of UX(0)∣BU_X(0)|_B is the line spanned by bb and (b,1)=H4=61(b,1)=H^4=61. As this line contains bχb_\chi and (bχ,1)=1(b_\chi,1)=1, we get bχ=b/61b_\chi=b/61.

(3) This is Corollary 4.3 since dim⁡VX=32\dim V_X=32. □

The following statement is used in Section 6.

Proposition 4.9. Let X⊂Y⊂GX\subset Y\subset G be a Küchle fourfold of type (c6) and P=c2(Q)∣XP=c_2(Q)|_X. Then the classes H2H^2 and PP generate a saturated sublattice of H4(X,Z)/torsH^4(X,\mathbb{Z})/\mathrm{tors}.

Proof. Let ω0,ω1∈⋀2(C7)∨\omega_0,\omega_1\in\bigwedge^2(\mathbb{C}^7)^\vee be the alternating 2-forms defining the smooth sixfold Y⊂GY\subset G, and let E⊂C7E\subset\mathbb{C}^7 be a common 4-dimensional isotropic subspace for ω0\omega_0 and ω1\omega_1 (any pair of alternating 2-forms on C2m+1\mathbb{C}^{2m+1} admits a common (m+1)(m+1)-dimensional isotropic subspace).

Choose w∈C7∖Ew\in\mathbb{C}^7\setminus E, a line ℓ⊂E\ell\subset E, and a 2-plane A⊂E∩w⊥ω0∩w⊥ω1A\subset E\cap w^{\perp_{\omega_0}}\cap w^{\perp_{\omega_1}}. Then the Schubert planes

Π0={L∈G∣ℓ⊂L⊂E}≅P2,Π1={L∈G∣A⊂L⊂E⊕⟨w⟩}≅P2\Pi_0=\{L\in G\mid\ell\subset L\subset E\}\cong\mathbb{P}^2,\qquad\Pi_1=\{L\in G\mid A\subset L\subset E\oplus\langle w\rangle\}\cong\mathbb{P}^2

lie in YY. Writing HY=c1(OY(1))H_Y=c_1(\mathcal{O}_Y(1)), standard Schubert calculus gives

(∫Π0HY2∫Π0c2(Q)∫Π1HY2∫Π1c2(Q))=(1011).\begin{pmatrix} \int_{\Pi_0}H_Y^2 & \int_{\Pi_0}c_2(Q) \\ \int_{\Pi_1}H_Y^2 & \int_{\Pi_1}c_2(Q) \end{pmatrix} = \begin{pmatrix} 1 & 0 \\ 1 & 1 \end{pmatrix}.

Because this pairing matrix is unimodular, HY2H^2_Y and c2(Q)∣Yc_2(\mathcal{Q})|_Y span a saturated rank-2 sublattice of H4(Y,Z)/torsH^4(Y,\mathbb{Z})/\mathrm{tors}. Since XX is cut out inside YY by two hyperplane sections, integral Weak Lefschetz shows that H4(Y,Z)↪H4(X,Z)H^4(Y,\mathbb{Z}) \hookrightarrow H^4(X,\mathbb{Z}) is injective with torsion-free cokernel, so saturation passes to H4(X,Z)/torsH^4(X,\mathbb{Z})/\mathrm{tors}.

Quantum Noether Rationality criterion

In [Gué26a, Proposition 26], Guéré proves an integral Hodge isometry between the transcendental part of a Hodge atom of a rational fourfold and the transcendental lattice of an associated minimal surface. Here we combine it with the Hodge atoms of [27] to prove Proposition 1.2. To do this, we consider the generalized eigenspaces of UX(τ)U_X(\tau) at general points of the non-archimedean Hodge locus of [27 Section 5.4].

We first recall the elementary lattice theory used below; see Nikulin [Nik80, Section 1].

Lattices

From now on, lattice means a free abelian group LL of finite rank equipped with a symmetric bilinear pairing (⋅,⋅):L×L⟶Z(\cdot,\cdot):L \times L \longrightarrow\mathbb{Z} that is nondegenerate over Q\mathbb{Q}. Write

L∨={x∈L⊗Q:(x,L)⊂Z},D(L)=L∨/L,L^\vee= \{x \in L \otimes\mathbb{Q} : (x,L) \subset\mathbb{Z}\}, \qquad D(L) = L^\vee/L,

and let ℓ(D(L))\ell(D(L)) denote the minimum number of generators of the discriminant group D(L)D(L). A lattice LL is unimodular if L=L∨L=L^\vee; the lattices ΛX\Lambda_X and ΛS\Lambda_S are unimodular by integral Poincaré duality. A sublattice L⊆ΛL \subseteq\Lambda is primitive (or saturated) if Λ/L\Lambda/L is torsion-free, equivalently if L=(L⊗Q)∩ΛL=(L\otimes\mathbb{Q})\cap\Lambda.

If ΓL\Gamma_L is the Gram matrix of LL in an integral basis, then

D(L)≅coker⁡ΓL,∣D(L)∣=∣det⁡ΓL∣,ℓ(D(L))≤rk⁡L.D(L) \cong\operatorname{coker}\Gamma_L, \qquad|D(L)|=|\det\Gamma_L|, \qquad\ell(D(L))\leq\operatorname{rk}L.

Thus the discriminant group can be computed from the Smith normal form of ΓL\Gamma_L.

If L⊂ML \subset M are lattices of the same rank, then M/LM/L is a finite group. Its number of elements, [M:L]:=∣M/L∣[M:L]:=|M/L|, is called the index of LL in MM. Since their Gram matrices satisfy

det⁡ΓL=[M:L]2det⁡ΓM,\det\Gamma_L=[M:L]^2\det\Gamma_M,

a sublattice is saturated whenever ∣det⁡ΓL∣|\det\Gamma_L| is squarefree.

If LL is primitive and nondegenerate and Λ\Lambda is unimodular, then D(L)≅D(L⊥)D(L)\cong D(L^\perp), as finite abelian groups [Nik80, Proposition 1.6.1].

Proof of Proposition 1.2. Suppose that XX is rational. By Guéré’s integral Hodge isometry [Gué26a, Propositions 25–26], applied to a weak factorization of a birational map X⇢P4X \dashrightarrow\mathbb{P}^4, there is a surface centre whose smooth minimal model SS satisfies

TX(1)≃TST_X(1)\simeq T_S

as integral Hodge structures. The isometry gives pg(S)=h3,1(X)>0p_g(S)=h^{3,1}(X)>0. Thus SS is not ruled and minimality implies that KSK_S is nef.

Since TX,QT_{X,\mathbb{Q}} is polarizable and End⁡Hdg(TX,Q)=Q⋅id⁡\operatorname{End}_{\mathrm{Hdg}}(T_{X,\mathbb{Q}})=\mathbb{Q}\cdot\operatorname{id}, it is a simple Hodge structure and it is not of Tate type (as h3,1(X)>0h^{3,1}(X)>0), where its orthogonal complement in Heven(X,Q)H^{\mathrm{even}}(X,\mathbb{Q}) is entirely Tate by (1.1). Given that Gromov–Witten classes are algebraic, it follows that UX(τ)U_X(\tau) preserves TX⊗KT_X\otimes K and acts on it by a scalar λX(τ)\lambda_X(\tau). In particular, TX,QT_{X,\mathbb{Q}} lives in a single Hodge atom of XX, which corresponds (under weak factorization) to an atom of SS and all of H∗(S,Q)H^\ast(S,\mathbb{Q}); see [27 Lemma 5.24]. After specialization this gives

r+ρ(S)+2+4h1,0(S)=dim⁡QH∗(S,Q)≤mX≤MX.(19)r+\rho(S)+2+4h^{1,0}(S)=\dim_{\mathbb{Q}}H^\ast(S,\mathbb{Q})\leq m_X\leq M_X. \tag*{(19)}

The lattices TST_S and NS⁡(S)/tors⁡\operatorname{NS}(S)/\operatorname{tors} are primitive orthogonal complements in the unimodular lattice ΛS\Lambda_S. Hence

D(TX)≃D(TS)≃D(NS⁡(S)/tors⁡),ℓ(D(TX))≤ρ(S).D(T_X) \simeq D(T_S) \simeq D(\operatorname{NS}(S)/\operatorname{tors}), \qquad\ell(D(T_X)) \leq\rho(S).

Together with ρ(S)≥1\rho(S) \geq1 and (19), this proves (1) and the lower bound in (2). Noether’s formula gives

KS2+ρ(S)+8h1,0(S)=10+12h3,1(X)−r.(20)K_S^2 + \rho(S) + 8h^{1,0}(S) = 10 + 12h^{3,1}(X) - r. \tag*{(20)}

Since KS2≥0K_S^2 \geq0, the upper bound in (2) follows.

Finally, suppose MX≤r+3M_X \leq r + 3. Equation (19) requires

ρ(S)=1,h1,0(S)=0,mX=MX=r+3.\rho(S) = 1, \qquad h^{1,0}(S) = 0, \qquad m_X = M_X = r + 3.

Consequently, e(S)=MXe(S) = M_X and

NS⁡(S)/tors⁡=Zu,u2=∣D(TX)∣,\operatorname{NS}(S)/\operatorname{tors} = \mathbb{Z}u, \qquad u^2 = \lvert D(T_X)\rvert,

so D(TX)D(T_X) is cyclic. Writing [KS]=ku[K_S] = ku modulo torsion, with k∈Zk \in\mathbb{Z}, Noether’s formula becomes

12(1+h3,1(X))=12χ(OS)=KS2+e(S)=∣D(TX)∣k2+MX.12(1 + h^{3,1}(X)) = 12\chi(\mathcal{O}_S) = K_S^2 + e(S) = \lvert D(T_X)\rvert k^2 + M_X.

This proves (3). □

Irrationality of Küchle fourfolds

Proposition 6.1. For a very general Küchle fourfold XX of type (b8), (b11), (c1) or (c6),

TX,Q=VXandEnd⁡Hdg(TX,Q)=Q⋅id⁡.T_{X,\mathbb{Q}} = V_X \qquad\text{and}\qquad\operatorname{End}_{\mathrm{Hdg}}(T_{X,\mathbb{Q}}) = \mathbb{Q}\cdot\operatorname{id}.

Moreover, with PP as in Section 4, a Z\mathbb{Z}-basis of AXA_X is given by (P,12c2(Q)∣X)(P,\frac{1}{2}c_2(Q)|_X) in type (b8), (eˉ1,eˉ2,eˉ3)(\bar e_1,\bar e_2,\bar e_3) in type (b11), and by (H2,P)(H^2,P) in types (c1) and (c6), with Gram matrices

Γc1=(42212111),Γc6=(61333318),Γb8=(8666),Γb11=(231393132).\Gamma_{\mathrm{c1}} = \begin{pmatrix} 42 & 21 \\ 21 & 11 \end{pmatrix}, \qquad \Gamma_{\mathrm{c6}} = \begin{pmatrix} 61 & 33 \\ 33 & 18 \end{pmatrix}, \qquad \Gamma_{\mathrm{b8}} = \begin{pmatrix} 8 & 6 \\ 6 & 6 \end{pmatrix}, \qquad \Gamma_{\mathrm{b11}} = \begin{pmatrix} 2 & 3 & 1 \\ 3 & 9 & 3 \\ 1 & 3 & 2 \end{pmatrix}.

In particular D(TX)≃D(AX)D(T_X) \simeq D(A_X) is isomorphic to Z/2⊕Z/6\mathbb{Z}/2 \oplus\mathbb{Z}/6 and (Z/3)2(\mathbb{Z}/3)^2, Z/21\mathbb{Z}/21, Z/9\mathbb{Z}/9, respectively.

Proof. For very general XX, the Hodge group of VXV_X contains the connected algebraic monodromy group SO⁡(VX)\operatorname{SO}(V_X); see [13 Proposition 7.5]. Its standard representation is absolutely irreducible and contains no nonzero invariant vectors. Thus VXV_X contains no nonzero rational Hodge classes and Schur’s lemma gives:

End⁡Hdg(VX)=Q⋅id⁡.\operatorname{End}_{\mathrm{Hdg}}(V_X) = \mathbb{Q}\cdot\operatorname{id}.

Since BX4B_X^4 is algebraic, it follows that

AX⊗ZQ=BX4,TX,Q=VX.A_X \otimes_{\mathbb{Z}} \mathbb{Q} = B_X^4, \qquad T_{X,\mathbb{Q}} = V_X.

Let LXL_X be the lattice generated by these classes, whose rational span is BX4B_X^4 by § 4.3.

  • For (c1), det⁡Γc1=21\det\Gamma_{\mathrm{c1}} = 21 is squarefree, so LXL_X is saturated.

  • For (b8), standard integral Schubert calculus on Y=OG⁡(2,7)Y = \operatorname{OG}(2,7) gives

H4(Y,Z)=ZHY2⊕Zτ2,c2(Q)∣Y=2τ2.H^4(Y,\mathbb{Z}) = \mathbb{Z}H_Y^2 \oplus\mathbb{Z}\tau_2, \qquad c_2(Q)|_Y = 2\tau_2.

Thus (c2(S∨)∣Y,τ2)(c_2(S^\vee)|_Y,\tau_2) is also an integral basis.

  • For (b11), the classes e1e_1, e2e_2, e3e_3 form an integral basis by Meng’s cellular calculation [36 Sections 3.3–4, Proposition 4.4].

  • For (c6), saturation follows from Proposition 4.9.

In types (b8) and (b11), integral weak Lefschetz through the successive smooth hyperplane sections shows that LX\mathcal{L}_X is primitive in ΛX\Lambda_X as well. Since its rational span is AX⊗ZQA_X \otimes_{\mathbb{Z}} \mathbb{Q}, this gives LX=AX\mathcal{L}_X=A_X.

The displayed Gram matrices are the same as those from §4, except for (b8), where H2=P+2θH^2=P+2\theta with θ=12c2(Q)∣X\theta=\frac{1}{2}c_2(\mathcal{Q})|_X and we use the basis (P,θ)(P,\theta). The stated discriminant groups follow from their Smith normal forms, and unimodularity of ΛX\Lambda_X identifies D(TX)D(T_X) with D(AX)D(A_X).

Proof of Theorem 1.1. Let XX be very general of one of the four types. By Proposition 6.1, XX satisfies (1.1) with TX,Q=VXT_{X,\mathbb{Q}}=V_X. The computations in § 4.3 show that we have MX=rk⁡TX+3M_X=\operatorname{rk}T_X+3 in each case.

Suppose that XX is rational. Proposition 1.2(3) then requires D(TX)D(T_X) to be cyclic. Hence types (b8) and (b11) cannot be rational by Proposition 6.1. Moreover, ∣D(TX)∣k2=9+12h3,1(X)−rk⁡TX\lvert D(T_X)\rvert k^2=9+12h^{3,1}(X)-\operatorname{rk}T_X for some k∈Zk\in\mathbb{Z}. For types (c1) and (c6) this is:

21k2=9,9k2=1,21k^2=9,\qquad9k^2=1,

respectively, and neither has an integer solution.

Applications to Verra fourfolds

We now apply Proposition 1.2 to recover Guéré’s result that the very general Verra fourfold is irrational [Gué26a, Corollary 28]. Recall that a Verra fourfold XX is a smooth double cover

π:X⟶P2×P2\pi:X\longrightarrow\mathbb{P}^2\times\mathbb{P}^2

branched over a divisor of bidegree (2,2)(2,2). It has two associated quadric surface fibrations which we denote pi=pr⁡i∘π:X⟶P2p_i=\operatorname{pr}_i\circ\pi:X\longrightarrow\mathbb{P}^2 for i=1,2i=1,2.

Proposition 7.1. For a very general Verra fourfold XX one has

TX(1)≅U⊕U(2)⊕E8(−1)⊕2⊕⟨−2⟩,D(TX)≅(Z/2Z)3,ℓ(D(TX))=3.T_X(1)\cong U\oplus U(2)\oplus E_8(-1)^{\oplus2}\oplus\langle-2\rangle,\qquad D(T_X)\cong(\mathbb{Z}/2\mathbb{Z})^3,\qquad\ell(D(T_X))=3.

TXT_X is primitive in ΛX\Lambda_X and has rk⁡TX=21\operatorname{rk}T_X=21. Furthermore End⁡Hdg(TX⊗ZQ)=Q⋅id⁡\operatorname{End}_{\mathrm{Hdg}}(T_X\otimes_{\mathbb{Z}}\mathbb{Q})=\mathbb{Q}\cdot\operatorname{id}.

Proof. TXT_X is primitive and nondegenerate by definition as AX≅U(2)⊕⟨2⟩A_X\cong U(2)\oplus\langle2\rangle is nondegenerate. The rank and discriminant group assertions follow from the stated lattice identification on the Hodge structure of TXT_X, which is provided in [IKKR17, Section 4].

Let SS be the degree two K3 surface associated with either quadric surface fibration pip_i. As rational Hodge structures we have

TX,Q(1)≅TS,Q:=TS⊗ZQ.T_{X,\mathbb{Q}}(1)\cong T_{S,\mathbb{Q}}:=T_S\otimes_{\mathbb{Z}}\mathbb{Q}.

As a consequence of [32 Proposition 2.10(ii)], the associated degree two K3 surfaces vary dominantly in their moduli space. Hence SS is very general when XX is very general. For such an SS, the Hodge group of T(S)QT(S)_{\mathbb{Q}} is SO⁡(T(S)Q)\operatorname{SO}(T(S)_{\mathbb{Q}}); see [21 Chapter 6, Corollary 4.7 and Theorem 4.9]. Thus End⁡Hdg(T(S)Q)=Q⋅id⁡\operatorname{End}_{\mathrm{Hdg}}(T(S)_{\mathbb{Q}})=\mathbb{Q}\cdot\operatorname{id}.

Proof of Corollary 1.3. By Proposition 7.1 a very general Verra fourfold XX satisfies (1.1) with rk⁡TX=21\operatorname{rk}T_X=21, h3,1(X)=1h^{3,1}(X)=1 and ℓ(D(TX))=3\ell(D(T_X))=3. If XX were rational, Proposition 1.2(2) would give 3≤ρ(S)≤10+12−21=13\leq\rho(S)\leq10+12-21=1.

Remark 7.2. Various works have compared Küchle fourfolds with Verra fourfolds. In particular, Benedetti–Fay–Guéré–Manivel–Perrin proved that Hodge-general Küchle fourfolds of type (c5) are not birational to any Verra fourfold [3 Corollary 5.3] and the author proved the same for very general Küchle fourfolds of type (c7) [15 Corollary 1.6]. Benedetti–Guéré–Manivel–Perrin studied the birational geometry of Verra fourfolds in [5]. For very general symmetric Verra fourfolds, equivariant irrationality with respect to the factor-swap involution was proved in [14 Theorem 6.4]. More recently, Jacovskis–Meinsma–Moschetti gave a construction of cubic fourfolds birational to Verra fourfolds in [25].

Code Availability

The code and data associated with the Quantum Lefschetz calculations in Section 4 can be found at https://github.com/aideenfay/kuchle-quantum-lefschetz. Our SageMath implementation of Atiyah–Bott equivariant localization for the line counts in Table 3 is available at https://github.com/aideenfay/grassmannian-lines.

Küchle typeGGF\mathcal{F}aXa_X
(b1)Gr⁡(2,5)\operatorname{Gr}(2,5)O(3)\mathcal{O}(3)18
(b5)Gr⁡(2,6)\operatorname{Gr}(2,6)S∨(1)⊕O(1)S^\vee(1) \oplus\mathcal{O}(1)7
(b6)Gr⁡(2,6)\operatorname{Gr}(2,6)O(1)⊕2⊕O(2)\mathcal{O}(1)^{\oplus2} \oplus\mathcal{O}(2)8
(b7)Gr⁡(2,7)\operatorname{Gr}(2,7)O(1)⊕5\mathcal{O}(1)^{\oplus5}5
(b8)Gr⁡(2,7)\operatorname{Gr}(2,7)Sym⁡2S∨⊕O(1)⊕2\operatorname{Sym}^2 S^\vee\oplus\mathcal{O}(1)^{\oplus2}4
(b11)Gr⁡(2,8)\operatorname{Gr}(2,8)Q∨(1)⊕O(1)Q^\vee(1) \oplus\mathcal{O}(1)3
(c1)Gr⁡(3,6)\operatorname{Gr}(3,6)O(1)⊕4\mathcal{O}(1)^{\oplus4}6
(c2)Gr⁡(3,6)\operatorname{Gr}(3,6)⋀2S∨⊕O(2)\bigwedge^2 S^\vee\oplus\mathcal{O}(2)8
(c4)Gr⁡(3,7)\operatorname{Gr}(3,7)Sym⁡2S∨⊕O(1)\operatorname{Sym}^2 S^\vee\oplus\mathcal{O}(1)0
(c5)Gr⁡(3,7)\operatorname{Gr}(3,7)⋀2S∨⊕Q∨(1)\bigwedge^2 S^\vee\oplus Q^\vee(1)4
(c6)Gr⁡(3,7)\operatorname{Gr}(3,7)(⋀2S∨)⊕2⊕O(1)(\bigwedge^2 S^\vee)^{\oplus2} \oplus\mathcal{O}(1)5
(c7)Gr⁡(3,8)\operatorname{Gr}(3,8)⋀3Q\bigwedge^3 Q0
(d1)Gr⁡(4,8)\operatorname{Gr}(4,8)Sym⁡2S∨⊕O(1)\operatorname{Sym}^2 S^\vee\oplus\mathcal{O}(1)0
(d3)Gr⁡(5,10)\operatorname{Gr}(5,10)(⋀2S∨)⊕2(\bigwedge^2 S^\vee)^{\oplus2}5

Table 3. The numbers aXa_X of lines through a general point of the fivefold Y′Y' defined by F\mathcal{F}, containing the fourfold XX as a hyperplane section.

Acknowledgments

The author thanks Tom Coates for his assistance and many valuable discussions, and gratefully acknowledges EPSRC grant EP/Y028872/1. Parts of this work were assisted by GPT-5.5 and GPT-6 Astra in the form of finding gaps in proofs and checking the final paper for mistakes.

References

Line counts by Atiyah–Bott Equivariant Localization

Let X⊂G=Gr⁡(k,n)X \subset G = \operatorname{Gr}(k,n) be a fourfold from Küchle’s list [Küc95] defined by the bundle F⊕O(1)\mathcal{F} \oplus\mathcal{O}(1), and let Y′Y' be the fivefold cut out by F\mathcal{F}. Thus XX is a hyperplane section of Y′Y' in the Plücker embedding G↪P(⋀kCn)G \hookrightarrow\mathbb{P}(\bigwedge^k \mathbb{C}^n). For a general point p∈Y′p \in Y', every line ℓ∈G\ell\in G through pp is parameterized by Pk−1×Pn−k−1\mathbb{P}^{k-1} \times\mathbb{P}^{n-k-1}, and those contained in Y′Y' are the zeros of a section of the bundle E\mathcal{E} with fibres

Eℓ=H0(ℓ,F∣ℓ(−p)).\mathcal{E}_{\ell} = H^0(\ell,\mathcal{F}|_{\ell}(-p)).

Since E\mathcal{E} has rank c1(F)⋅ℓ=n−2c_1(\mathcal{F}) \cdot\ell= n-2, the number of lines through pp is

aX=∫Pk−1×Pn−k−1cn−2(E).a_X = \int_{\mathbb{P}^{k-1} \times\mathbb{P}^{n-k-1}} c_{n-2}(\mathcal{E}).

We compute this integral using Atiyah–Bott equivariant localization [1 Equation 3.8] and record the results in Table 3. See § 8 for information on code availability.

Department of Mathematics, Imperial College London, London SW7 2AZ, United Kingdom

Email address: [email protected]

References

  1. [1][AB84] M.F. Atiyah and R. Bott. The moment map and equivariant cohomology. Topology, 23(1):1–28, 1984.
  2. [2][ACG+26] Boris Alexeev, Leonardo F. Cavenaghi, Giovane Galindo, Bogdan Georgiev, Ludmil Katzarkov, and Pedro Antonio Muniz Martins. On the atomic decomposition of complete intersection in flag varieties, 2026.arxiv.org/abs/2609.21914
  3. [3][BFG+26] Vladimiro Benedetti, Aideen Fay, Jérémy Guéré, Laurent Manivel, and Nicolas Perrin. An atomic criterion for irrationality without quantum computations, 2026.arxiv.org/abs/2607.26718
  4. [4][BFMT26] Marcello Bernardara, Enrico Fatighenti, Laurent Manivel, and Fabio Tanturri. Fano fourfolds of K3 type, 2026.arxiv.org/abs/2111.13030
  5. [5][BGMP26] Vladimiro Benedetti, Jérémy Guéré, Laurent Manivel, and Nicolas Perrin. Quantum cohomology and birational geometry of Verra fourfolds, 2026.arxiv.org/abs/2605.30450
  6. [7][CG07] Tom Coates and Alexander B. Givental. Quantum Riemann–Roch, Lefschetz and Serre. Annals of Mathematics, 165:15–53, 2007.DOI
  7. [8][CGKS20] Tom Coates, Sergey Galkin, Alexander Kasprzyk, and Andrew Strangeway. Quantum periods for certain four-dimensional Fano manifolds. Experimental Mathematics, 29(2):183–221, 2020.
  8. [9][CJ99] Alberto Collino and Masao Jinzenji. On the structure of the small quantum cohomology rings of projective hypersurfaces. Communications in Mathematical Physics, 206(1):157–183, September 1999.
  9. [10][Coa14] Tom Coates. The Quantum Lefschetz principle for vector bundles as a map between Givental cones, 2014.arxiv.org/abs/1405.2893
  10. [11]Pierre Deligne. La conjecture de Weil. I. Publications Mathématiques de l’Institut des Hautes Études Scientifiques, 43(1):273–307, 1974.DOI
  11. [12]Pierre Deligne. La conjecture de Weil : II. Publications Mathématiques de l’IHÉS, 52:137–252, 1980.DOI
  12. [13]Pierre Deligne. La conjecture de Weil pour les surfaces K3. Inventiones mathematicae, 15:206–226, 1971/72.DOI
  13. [14]Aideen Fay. Equivariant irrationality of very general symmetric Verra fourfolds, 2026.arxiv.org/abs/2605.30439
  14. [15]Aideen Fay. Hasse obstructions to rationality for special fourfolds, 2026.arxiv.org/abs/2609.06759
  15. [16]William Fulton. Intersection Theory, volume 2 of Ergebnisse der Mathematik und ihrer Grenzgebiete. 3. Folge / A Series of Modern Surveys in Mathematics. Springer-Verlag, Berlin, Heidelberg, 1984.
  16. [17]V. Golyshev. Spectra and strains, 2008.arxiv.org/abs/0801.0432
  17. [20]Xiaowen Hu. Big quantum cohomology of Fano complete intersections, 2025.
  18. [21]Daniel Huybrechts. Lectures on K3 Surfaces. Cambridge Studies in Advanced Mathematics. Cambridge University Press, 2016.DOI
  19. [22]Daisuke Inoue, Atsushi Ito, and Makoto Miura. I-functions of Calabi–Yau 3-folds in Grassmannians. Communications in Number Theory and Physics, 11(2):273–309, 2017.DOI
  20. [24]Hiroshi Iritani. Quantum ring and Quantum Lefschetz, 2017.
  21. [25]Augustinas Jacovskis, Reinder Meinsma, and Riccardo Moschetti. The Verra involution on special cubic fourfolds, 2026.arxiv.org/abs/2610.32952
  22. [26]Nicholas M. Katz. Larsen’s alternative, moments, and the monodromy of Lefschetz pencils. In Contributions to automorphic forms, geometry, and number theory, pages 521–560. Johns Hopkins University Press, Baltimore, MD, 2004.
  23. [27]Ludmil Katzarkov, Maxim Kontsevich, Tony Pantev, and Tony Yue Yu. Birational invariants from Hodge structures and quantum multiplication, 2026.arxiv.org/abs/2508.05105
  24. [28]M. Kontsevich and Yu. Manin. Gromov–Witten classes, quantum cohomology, and enumerative geometry. Communications in Mathematical Physics, 164(3):525–562, August 1994.
  25. [30]A G Kuznetsov. On Küchle varieties with Picard number greater than 1. Izvestiya: Mathematics, 79(4):698–709, August 2015.DOI
  26. [31]Alexander Kuznetsov. Küchle fivefolds of type c5. Mathematische Zeitschrift, 284(3–4):1245–1278, 2016.DOI
  27. [32]Robert Laterveer. Algebraic cycles and Verra fourfolds. Tohoku Mathematical Journal, Second Series, 72(3):451–485, 2020.
  28. [33]James D. Lewis. The cylinder homomorphism associated to quintic fourfolds. Compositio Mathematica, 56:315–329, 1985.
  29. [34]James D. Lewis. The cylinder correspondence for hypersurfaces of degree n in ℙⁿ. American Journal of Mathematics, 110(1):77–114, 1988.DOI
  30. [35]Jun Li and Gang Tian. Virtual moduli cycles and Gromov–Witten invariants of algebraic varieties, 1998.arxiv.org/abs/alg-geom/9602007
  31. [36]Junyu Meng. Quantum cohomology of a Fano quiver moduli space, 2024.arxiv.org/abs/2412.15987
  32. [38]Ichiro Shimada. On the cylinder homomorphisms of Fano complete intersections. Journal of the Mathematical Society of Japan, 42(4):719 – 738, 1990.DOI

Paper details

vlnc:W_39m93gee
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

GPT-5.5 (bare), GPT-6 Astra

View statement
Looking for related papers…
Contents