🔢 mathematics

Fourier-Mukai partners of non-syzygetic cubic fourfolds and Gale duality

This paper establishes that very general non-syzygetic cubic fourfolds possess a unique nontrivial Fourier-Mukai partner which is also non-syzygetic and obtained via Gale duality, while demonstrating that although their Fano varieties of lines remain birational, the cubics themselves may not, thereby providing potential counterexamples to conjectures by Brooke-Frei-Marquand and Huybrechts regarding equivariant birationality.

Original authors: Christian Böhning, Hans-Christian Graf von Bothmer, Lisa Marquand

Published 2026-08-13

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Christian Böhning, Hans-Christian Graf von Bothmer, Lisa Marquand

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are exploring a vast, invisible landscape made not of mountains and rivers, but of pure mathematical shapes called "four-dimensional cubes." In the world of algebraic geometry, these aren't the dice you roll in a board game; they are complex, smooth surfaces defined by equations, living in a space so high-dimensional that our brains can't easily picture them. Mathematicians are obsessed with figuring out when two of these shapes are actually the same thing, just dressed up differently. Sometimes, two shapes look totally different on the outside, but if you peel back the layers of their internal structure, you find they are secretly twins. This is the heart of a big question in modern math: If two shapes are "Fourier-Mukai partners" (a fancy way of saying their internal mathematical DNA is identical), does that mean you can stretch and squish one into the other without tearing it apart? This is called being "birational." It's like asking if a lump of clay shaped like a cat can be reshaped into a dog without adding or removing any clay.

For a long time, mathematicians suspected the answer was always "yes." They thought that if the internal DNA matched, the shapes must be transformable into each other. But recently, a team of researchers decided to test this rule using a very specific, tricky type of four-dimensional cube called a "non-syzygetic cubic fourfold." These are special shapes that contain two smaller, scroll-like surfaces that intersect in a very precise way. The researchers wanted to see if the "DNA match" rule held up when these shapes were twisted by symmetry groups, like rotating a snowflake. They built a new machine to generate these shapes and their "partners," hoping to find a case where the DNA matched perfectly, but the shapes refused to turn into one another.

In this paper, Christian Böhm, Hans-Christian Graf von Bothmer, and Lisa Marquand take a deep dive into these special cubes. They prove that for a "very general" (meaning a typical, random) non-syzygetic cubic fourfold, there is exactly one other shape that shares its internal DNA. They call this partner the "Gale dual." Think of it like a mirror image, but instead of flipping left and right, the mirror flips the mathematical ingredients of the equation in a way that comes from a branch of math called convex geometry. The authors show that you can take the equation of the first shape, run it through a specific algebraic recipe (involving a matrix and some linear forms), and out pops the equation of its partner.

The big discovery here is a bit of a twist. The authors prove that these two partner shapes are indeed "Fourier-Mukai partners"—their internal structures are identical. They also prove that their "Fano varieties of lines" (which are collections of all the straight lines that can be drawn inside the shape) are birational, meaning those collections of lines can be transformed into each other. However, the paper leaves a door open for a potential counterexample to the big "DNA means shape" rule. When the researchers added a layer of symmetry (like asking the shapes to respect the rotations of a tetrahedron), they found that while the line collections still matched up perfectly under the symmetry, the shapes themselves might not be transformable into each other in a way that respects that symmetry.

Specifically, the authors construct examples using the alternating group on four letters (a specific symmetry group) where the partner shapes have faithful actions of this group. In these cases, the Fano varieties of lines remain "symmetric-birational," but the authors could not find a way to prove the shapes themselves are symmetric-birational. They suggest these examples might be the first cracks in the wall of a famous conjecture by Daniel Huybrechts, which predicts that matching DNA always means the shapes are transformable. While they haven't definitively proven the conjecture false yet, they have built a very strong candidate for a counterexample, showing that in the world of symmetric shapes, having the same internal code doesn't guarantee you can morph one into the other without breaking the rules of the symmetry.

Technical Summary: Fourier-Mukai Partners of Non-Syzygetic Cubic Fourfolds and Gale Duality

Problem Statement
The paper addresses the classification of smooth cubic fourfolds and their birational types, specifically focusing on the relationship between derived categories and birational geometry. The central problem is to understand the structure of Fourier-Mukai (FM) partners—smooth cubic fourfolds $X$ and $X'$ whose Kuznetsov components $A_X$ and $A_{X'}$ are equivalent—and to determine when such equivalence implies birationality. This relates to Huybrechts' conjecture, which posits that if $A_X \simeq A_{X'}$ , then $X$ and $X'$ are birational. The authors focus on a specific class of cubic fourfolds known as "non-syzygetic" cubic fourfolds, defined as those containing two cubic surface scrolls in distinct hyperplanes with an intersection number of 1. Previous work by Fan and Lai and by Brooke, Frei, and Marquand (BFM) established the existence of such partners for very general non-syzygetic cubics but relied on transcendental techniques (Torelli theorems) that did not yield explicit equations for the partners.

Methodology
The authors employ a combination of multilinear algebra, commutative algebra, and representation theory to construct these FM partners explicitly.

Algebraic Characterization: They first characterize non-syzygetic cubic fourfolds algebraically. They prove that any such fourfold $X$ admits an equation of the form $\det(M) + L_1L_2L_3 = 0$ , where $M$ is a $3 \times 3$ matrix of linear forms and $L_i$ are linear forms.
Gale Duality: The core construction involves applying Gale duality to the linear forms defining $X$ . By treating the coefficients of the matrix $M$ and the linear forms $L_i$ as a linear map $\mathbb{C}^{12} \to \mathbb{C}^6$ , they define the kernel as a dual space of 12 linear forms. This yields a dual matrix $M'$ and dual forms $L'_i$ , defining a new cubic $X'$ via the equation $\det(M') - L'_1L'_2L'_3 = 0$ .
$\rho$ -Lagrangian Data: The authors introduce a linear algebraic datum called a " $\rho$ -Lagrangian data set" $(A, E, F)$ , where $A \subset \Lambda^3(E \oplus F)$ is a specific Lagrangian subspace. They establish a bijection between pairs of Gale dual cubics and these data sets.
Connection to Gushel-Mukai Fourfolds: They link these data sets to the geometry of Gushel-Mukai (GM) fourfolds. Specifically, they show that the $\rho$ -Lagrangian data parametrizes EPW sextics containing two disjoint rational surfaces isomorphic to $\mathbb{P}^2$ . Points on these surfaces correspond to GM fourfolds containing $\rho$ -planes (sub-Grassmannians $\text{Gr}(2, V_3)$ ).
Equivariant Analysis: The linear algebraic nature of the construction allows the authors to study these objects under the action of finite groups $G$ . They investigate whether the birational maps and FM equivalences hold $G$ -equivariantly.

Key Contributions and Results

Explicit Construction of FM Partners: The primary result is an explicit algebraic method to write down the equation of the unique nontrivial Fourier-Mukai partner $X'$ of a very general non-syzygetic cubic fourfold $X$ . This is achieved via the Gale dual construction, bypassing the need for transcendental Torelli techniques.
Uniqueness: They prove that a very general non-syzygetic cubic fourfold has precisely one nontrivial Fourier-Mukai partner. This is established by counting the number of overlattices in the primitive cohomology lattice, adapting methods from Fan and Lai.
Birationality of Partners: The paper proves that the Gale dual cubic $X'$ is birational to $X$ (though not isomorphic). This is shown by demonstrating that both $X$ and $X'$ can be obtained by projecting Gushel-Mukai fourfolds (which are period partners) from specific $\rho$ -planes.
Fano Varieties of Lines: The authors show that the Fano varieties of lines $F(X)$ and $F(X')$ are birational. This result is derived explicitly from the $\rho$ -Lagrangian data, recovering and refining previous results by BFM.
Equivariant Counterexamples: The paper constructs examples of Gale dual cubics with faithful actions of the alternating group $A_4$ . In these cases, while the Fano varieties of lines remain $A_4$ -birational, the cubics $X$ and $X'$ may not be $A_4$ -birational. This is because the intermediate Gushel-Mukai fourfolds required to establish the birational map may not exist as $G$ -varieties.
Conjectural Implications: The authors exhibit these $A_4$ $A_{4}$ -examples as potential counterexamples to the equivariant versions of:
- Conjecture 1.3 (BFM): If $F(X)$ and $F(X')$ are $G$ -birational, then $X$ and $X'$ are $G$ -birational.
- Conjecture 1.1 (Huybrechts): If $X$ and $X'$ are $G$ -Fourier-Mukai partners, then they are $G$ -birational.

Significance
The paper provides a concrete, algebraic realization of Fourier-Mukai partners for a specific class of cubic fourfolds, moving beyond abstract existence proofs. By linking Gale duality to the geometry of GM fourfolds and EPW sextics, it offers a new perspective on the birational geometry of these varieties. The explicit nature of the construction is highlighted as a tool for enumerating FM partners (as seen in related work enumerating 1120 partners for cubics with involutions). Furthermore, the equivariant analysis challenges the universality of the birationality conjectures in the presence of group actions, suggesting that the relationship between derived equivalence and birationality is more subtle in the equivariant setting. The work recovers and makes precise the results of BFM while opening new avenues for testing conjectures regarding equivariant birational geometry.

More like this