---
title: Peskine Sixfolds in Pfaffian Geometry
url: https://www.emergentmind.com/topics/peskine-sixfolds
type: topic
---

# Peskine Sixfolds in Pfaffian Geometry

Peskine sixfolds are six-dimensional Pfaffian degeneracy loci in $\mathbf P^9$ attached to a trivector $\sigma \in \bigwedge^3 V_{10}^\vee$, where $V_{10}$ is a $10$-dimensional complex vector space. In the current Debarre–Voisin literature, a Peskine sixfold is the variety
$$
X_1^\sigma := \{[\ell]\in \mathbf P(V_{10}) \mid \operatorname{rank}\sigma(\ell,-,-)\le 6\},
$$
which is smooth of dimension $6$ for general $\sigma$ and is birationally governed by the same Hodge-theoretic structures that control the associated Debarre–Voisin hyperkähler fourfold $X_6^\sigma$ [2211.16129; 2509.18339]. The subject lies at the intersection of skew-determinantal geometry, period maps, rationality problems, and the study of special loci carrying associated K3 surfaces or cubic fourfolds.

## 1. Definition and projective-geometric realization

Let $V$ be a $10$-dimensional complex vector space and $\sigma \in \bigwedge^3 V^\vee$ a nonzero trivector. One description of the associated Peskine sixfold takes $\sigma$ as a section of $\bigwedge^2 Q(1)$ on $\mathbf P(V)=\mathbf P^9$, where $Q$ is the quotient bundle, and defines
$$
X_\sigma := \{[\ell]\in \mathbf P(V)\mid \operatorname{rank}(\sigma|_{\ell,V})\le 6\}.
$$
Equivalently, a general trivector defines a skew-symmetric $9\times 9$ matrix of linear forms on $\mathbf P^9$ whose Pfaffian is a cubic hypersurface, and $X_\sigma$ is the singular locus of that cubic [2211.16129].

A complementary description uses the contraction map
$$
\sigma_v: V_{10}\to \bigwedge^2 V_{10}^\vee,\qquad
w\mapsto [(x,y)\mapsto \sigma(v,x,y)].
$$
In homogeneous coordinates on $\mathbf P(V_{10})\cong \mathbf P^9$, this becomes a $10\times 10$ skew-symmetric matrix of linear forms $M(v)$, and one sets
$$
X_1^\sigma=\{[v]\in \mathbf P^9\mid \operatorname{rank}(\sigma_v)\le 6\}.
$$
Since the expected codimension of the rank-$\le 6$ locus for a $10\times 10$ skew-symmetric matrix is $3$, one obtains a six-dimensional projective variety [2106.13287].

For a smooth general member, the sixfold is a linear Pfaffian variety of codimension $3$ and degree $15$, cut out by the vanishing of all $8\times 8$ Pfaffians of $M(v)$. There are $\binom{10}{8}=45$ such quartic equations [2106.13287; 2501.16013]. The rationality paper further records that one checks
$$
K_{X_\sigma}=\mathcal O_{X_\sigma}(-3),
$$
placing the general object in the Fano range [2211.16129].

The terminology is not completely uniform across the literature. In one unrelated source, the label “Peskine six-fold” is applied to the orthogonal Grassmannian $\mathrm{OG}^+(4,8)$ parametrizing $\mathbf P^3$’s on a smooth $6$-quadric [1111.3382]. In the recent trivector-based literature, however, “Peskine sixfold” refers to the six-dimensional degeneracy locus $X_1^\sigma \subset \mathbf P^9$ associated with a trivector in ten variables [2211.16129; 2509.18339].

## 2. Cohomology, algebraic cycles, and the Debarre–Voisin correspondence

For a smooth Peskine sixfold $X_1^\sigma$, the cohomology is organized by two distinguished algebraic classes in $H^{3,3}(X_1^\sigma,\mathbf Z)$: the cube of the hyperplane class,
$$
h^3=H^3,
$$
and the class $\pi$ of a Palatini threefold, namely a linear section $X_1^\sigma\cap \mathbf P(V_6)$ for $[V_6]\in X_6^\sigma$ [2509.18339]. By Benedetti–Song’s theorem as quoted there, the integral Hodge conjecture holds on $X_1^\sigma$, so every class in $H^{3,3}(X_1^\sigma,\mathbf Z)$ is algebraic [2509.18339].

The classes $h^3$ and $\pi$ span a rank-$2$ sublattice
$$
\Lambda_{11}:=\operatorname{Span}\{h^3,\pi\}\subset H^6(X_1^\sigma,\mathbf Z)
$$
with intersection matrix
$$
\langle h^3,h^3\rangle = 15,\qquad
\langle h^3,\pi\rangle = 7,\qquad
\langle \pi,\pi\rangle = 4.
$$
The orthogonal complement
$$
H^6_{\mathrm{van}}(X_1^\sigma):=(\Lambda_{11})^\perp
$$
is a unimodular lattice of signature $(2,20)$ carrying a polarized Hodge structure of K3 type of discriminant $11$ [2509.18339].

The key bridge to hyperkähler geometry is the Debarre–Voisin fourfold
$$
X_6^\sigma:=\{[V_6]\in \operatorname{Gr}(6,V_{10})\mid \sigma|_{V_6}=0\}.
$$
When both $X_1^\sigma$ and $X_6^\sigma$ are smooth, the incidence correspondence induces an integral Hodge isometry
$$
\varphi: H^6_{\mathrm{van}}(X_1^\sigma,\mathbf Z)(+1)\xrightarrow{\sim} H^2_{\mathrm{prim}}(X_6^\sigma,\mathbf Z),
$$
or equivalently
$$
\varphi: H^6_{\mathrm{van}}(X_1^\sigma,\mathbf Z)\xrightarrow{\sim} H^2_{\mathrm{prim}}(X_6^\sigma,\mathbf Z)(-1)
$$
depending on sign convention [2509.18339]. This identification is central: extra algebraic classes on the sixfold correspond to special algebraic classes on the Debarre–Voisin fourfold, and the period theory of one feeds directly into the other.

This parallelism is one reason the literature treats Peskine sixfolds as varieties “of K3 type.” The phrase is not merely heuristic: it refers to the existence of a K3-type polarized Hodge structure in middle cohomology and its integral linkage to the primitive $H^2$ of a hyperkähler fourfold [2509.18339].

## 3. Moduli and special loci

Up to the natural $\mathrm{SL}(V)$-action, general trivectors determine a $20$-dimensional GIT quotient
$$
M \simeq \mathbf P(\bigwedge^3 V^\vee)\!/\!/\mathrm{SL}(V),
$$
which is birational to the moduli space of Peskine sixfolds and of the associated Debarre–Voisin hyperkähler fourfolds [2211.16129]. Within this moduli space one studies special divisors arising from extra algebraic cycles, or equivalently from special lattice embeddings in the associated Hodge structures.

On the Debarre–Voisin side, the relevant period domains carry Heegner divisors $D_d$ indexed by positive even $d \equiv 0,2,6,8,10,18 \pmod{22}$, and their preimages in the GIT moduli of trivectors are denoted
$$
P_d \subset \mathbf P(\bigwedge^3 V_{10}^\vee)\!/\!/\mathrm{SL}(V_{10}).
$$
A trivector $\sigma\in P_d$ is called special of discriminant $d$ [2509.18339]. For a general point of $P_d$, the sixfold and fourfold carry marked algebraic lattices $M_d$ and $N_d$ of discriminant $d$ extending the basic lattices $\Lambda_{11}$ and the polarization class on $X_6^\sigma$ [2509.18339].

Two special divisors play a particularly prominent role in the rationality theory. In the notation of Benedetti–Song, one defines
$$
D^{3,3,10}=\{\sigma\in \bigwedge^3 V^\vee \mid \exists \text{ a $3$-plane }\Pi\subset V \text{ with } \sigma|_{\Pi,\Pi,V}\equiv 0\},
$$
and
$$
D^{1,6,10}=\{\sigma\in \bigwedge^3 V^\vee \mid \exists \text{ a flag }L\subset W\subset V,\ \dim L=1,\ \dim W=6,\ \sigma(L,W,W)=0\}.
$$
These are $\mathrm{SL}(V)$-invariant divisors in the trivector space [2211.16129].

The same moduli framework also organizes the existence of associated K3 surfaces and cubic fourfolds. A polarized K3 surface $(S,f)$ of degree $d$ is associated to $\sigma\in P_d\setminus P_{22}$ if there is a Hodge isometry
$$
N_d^\perp \subset H^2(X_6^\sigma,\mathbf Z)\cong f^\perp \subset H^2(S,\mathbf Z),
$$
and, away from $P_{22}\cup P_{24}$, equivalently an isometry
$$
M_d^\perp \subset H^6(X_1^\sigma,\mathbf Z)\cong f^\perp(-1)\subset H^2(S,\mathbf Z)(-1).
$$
Similarly, a marked cubic fourfold $(Y,K_d)$ of discriminant $d$ is associated to $\sigma$ if
$$
N_d^\perp \subset H^2(X_6^\sigma,\mathbf Z)\cong K_d^\perp(-1)\subset H^4(Y,\mathbf Z)(-1)
$$
[2509.18339].

## 4. Rationality and the cubic-fourfold analogy

The principal rationality result currently available is the theorem that if $\sigma\in D^{3,3,10}$ is general, then $X_\sigma$ is rational [2211.16129]. The proof fixes a $3$-subspace $\Pi\subset V$ with $\sigma(\Pi,\Pi,V)=0$ and considers projection from $\mathbf P(\Pi)$,
$$
X_\sigma \dashrightarrow \mathbf P(V/\Pi)=\mathbf P^6.
$$
A dimension count together with the exclusion of higher-dimensional fibers shows that this rational map is generically one-to-one, hence birational [2211.16129].

For $\sigma\in D^{1,6,10}$, the geometry is subtler. The sixfold acquires a natural divisor $D\subset X_\sigma$ ruled in conics, and there is a $\mathbf P^1$-fibration
$$
D \to S
$$
over a smooth K3 surface $S$ of degree $6$, inducing a Brauer class $\beta$ on $S$ [2211.16129]. In this setting the paper proves a cohomological criterion: if there exists a $4$-dimensional algebraic cycle $Z\subset X_\sigma$ such that
$$
(Z\cdot H\cdot h)\ \text{is odd},
$$
where $H$ is the hyperplane class on $X_\sigma\subset \mathbf P^9$ and $h$ is its pullback from the $\mathbf P^6$-bundle model, then the Brauer twist $\beta$ is trivial and $X_\sigma$ is rational [2211.16129].

The significance of this criterion is Hodge-theoretic. The existence of such a cycle corresponds to an extra integral Hodge class in $H^4(X_\sigma,\mathbf Z)$ of type $(2,2)$ splitting off the “K3” summand, the same summand that is isomorphic to the primitive Hodge structure of the associated Debarre–Voisin fourfold [2211.16129]. The mechanism is explicitly modeled on the classical rationality story for cubic fourfolds containing a plane: in both settings, a conic fibration produces a twisted K3 surface, and odd intersection conditions detect the triviality of the Brauer class.

This parallel is not merely formal. The paper conjectures that for general $\sigma\in D^{1,6,10}$ with associated twisted K3 surface $(S,\beta)$ of degree $6$, one has
$$
\beta=0 \quad \Longleftrightarrow \quad X_\sigma\ \text{is rational}.
$$
One direction is proved, namely odd-intersection $4$-cycles $\Rightarrow \beta=0 \Rightarrow X_\sigma$ rational [2211.16129]. It is further proposed that the conjecture should admit a derived-categorical reformulation: the Kuznetsov component of $D^b(X_\sigma)$ should become equivalent to $D^b(S)$ precisely when $\beta=0$ [2211.16129]. This suggests that Peskine rationality is controlled by the same mixture of lattice theory, twisted K3 geometry, and birational models that governs special cubic fourfolds.

## 5. Associated K3 surfaces and associated cubic fourfolds

The existence of associated K3 surfaces is governed by an explicit numerical criterion. For a very general $\sigma\in P_d$, there exists an associated K3 surface of degree $d$ if and only if
- $4\nmid d$ and $121\nmid d$;
- every odd prime dividing $d$ is a square mod $11$ [2509.18339].

The existence of associated cubic fourfolds is also characterized arithmetically. A very general $\sigma\in P_d$ admits an associated cubic fourfold of discriminant $d$ if and only if
- $d\equiv 0$ or $2 \pmod 6$;
- $9\nmid d$ and $121\nmid d$;
- $33$ is a square mod each prime divisor of $d$;
- if $66\mid d$, then $8\mid d$ and an odd number of primes $\equiv 2 \pmod 3$ divide $d$ [2509.18339].

The first discriminant for which an associated cubic appears is $d=24$ [2509.18339]. In that case, for a general $\sigma\in P_{24}$ with its unique flag
$$
W_1\subset W_6\subset V_{10},
$$
of dimensions $1$ and $6$ satisfying $\sigma(W_1,W_6,V_{10})=0$, the linear section
$$
Y=X_1^\sigma\cap \mathbf P(W_6)
$$
is a smooth cubic fourfold of discriminant $24$ [2509.18339]. The same paper proves that for very general $\sigma\in P_{24}$:
1. $Y$ is a smooth cubic fourfold of discriminant $24$;
2. the Fano variety of lines $F(Y)$ is isomorphic as a hyperkähler manifold to the associated Debarre–Voisin fourfold $X_6^\sigma$;
3. $Y$ is Hodge-theoretically associated to $\sigma$ [2509.18339].

This discriminant-$24$ case makes the Peskine–Debarre–Voisin/cubic-fourfold analogy fully geometric. The lines on $X_1^\sigma$ are controlled by the Grassmannian degeneracy locus
$$
X_7^\sigma=\{[V_7]\in \operatorname{Gr}(7,V_{10})\mid \operatorname{rank}\sigma|_{V_7}\le 5\},
$$
and for general $\sigma\in P_{24}$ there is a bijective morphism
$$
\phi: X_7^\sigma \to F(X_1^\sigma)
$$
to the Hilbert scheme of lines on the Peskine sixfold [2509.18339]. The induced map
$$
\psi:F(Y)\to X_6^\sigma
$$
is then shown to be an isomorphism.

The broader implication is that special Peskine sixfolds can carry cubic-fourfold data in precisely the same Hodge-theoretic sense in which special cubic fourfolds can carry K3 data. This suggests a bidirectional dictionary among Peskine sixfolds, Debarre–Voisin fourfolds, K3 surfaces, and special cubics, organized by discriminant and lattice-theoretic conditions [2509.18339].

## 6. Distinguished examples and auxiliary constructions

A particularly symmetric example comes from the finite simple group $\mathrm{PSL}(2,\mathbf F_{11})$. Let $V_{10}$ be its irreducible $10$-dimensional representation and $\sigma\in \bigwedge^3 V_{10}^\vee$ the unique nonzero $\mathrm{PSL}(2,\mathbf F_{11})$-invariant trivector. The associated Debarre–Voisin fourfold $X_6^\sigma$ is smooth, while the associated Peskine variety $X_1^\sigma\subset \mathbf P^9$ is highly symmetric and has exactly $55$ isolated singular points [2106.13287]. In this example the singular locus is the rank-$\le 4$ locus; the rank-$\le 2$ locus is empty; each of the $55$ reduced points is an ordinary double point; and the group acts transitively on the $55$ nodes [2106.13287]. The example serves as a controlled singular degeneration of the general smooth picture and illustrates how symmetries can force unexpectedly large special lattices on the Debarre–Voisin side.

Another construction arises from Mukai’s geometry of genus-$16$ K3 surfaces. For a general polarized K3 surface $(S,h)$ of genus $16$, Mukai’s rank-$2$ bundle $F$ gives a threefold
$$
X=\mathbf P_S(F)\hookrightarrow \mathbf P^9
$$
whose ideal is generated by $10$ independent quadrics [2501.16013]. The syzygies among these quadrics produce a canonical trivector
$$
\sigma=t_2\in \bigwedge^3 V_{10},
$$
hence a Peskine sixfold
$$
X_\sigma=\{[x]\in \mathbf P(V_{10})\mid \operatorname{rank}M(x)\le 6\}\subset \mathbf P^9,
$$
which is smooth, irreducible, and cut out by the $45$ quartic Pfaffians of the associated skew matrix [2501.16013].

In this genus-$16$ setting, the Peskine sixfold governs the failure of a double cover to remain finite. Since the ideal of the threefold $X$ is generated by the $10$-dimensional space of quadrics $V_{10}$, the linear system $|I_X(2)|$ defines, after blowing up $X$, a $2{:}1$ morphism
$$
f_X:\operatorname{Bl}_X\mathbf P^9\to \mathbf P(V_{10}^*)\simeq \mathbf P^9
$$
branched along a hypersurface of degree $10$ [2501.16013]. Proposition 3.21 identifies the associated Peskine sixfold as the locus over which this map ceases to be finite of degree $2$: for a general point $[p]\in X_\sigma$, the fiber $f_X^{-1}(p)$ is a single smooth rational cubic $C\subset X$, rather than two points [2501.16013].

Over such points one obtains further surface geometry. The restriction of $f_X$ to the $3$-plane $\mathbf P(E_{\sigma,p})$ is again a double cover branched along the union of a smooth cubic surface $S_{3,p}$, called the Weddle, and a Kummer quartic $K_p$ with $16$ nodes [2501.16013]. This yields a relative Weddle–Kummer surface fibration over the Peskine sixfold. The same paper shows that the kernel bundle $E_\sigma$ is isotropic for the trivector and defines a morphism
$$
X_\sigma \to DV(\sigma)\subset G(6,V_{10}),
$$
placing the sixfold inside the same hyperkähler network as the Debarre–Voisin variety [2501.16013].

Taken together, these constructions show that Peskine sixfolds occupy a central position in several parallel theories: skew-Pfaffian Fano geometry, period maps for hyperkähler fourfolds, special-lattice phenomena in moduli, rationality problems modeled on cubic fourfolds, and geometric constructions from polarized K3 surfaces. The current literature treats them not as isolated degeneracy loci, but as nodal points in a larger web linking Hodge theory, birational geometry, and explicit projective models [2211.16129; 2509.18339].

Source: https://www.emergentmind.com/topics/peskine-sixfolds