🔢 mathematics

Unpolarized Shafarevich conjectures for hyper-Kähler varieties

This paper proves the unpolarized Shafarevich conjecture for hyper-Kähler varieties within a given deformation type by unifying previous results on K3 surfaces and polarized hyper-Kähler varieties through the development of a uniform Kuga–Satake map, while also establishing finiteness results for varieties of CM type and discussing cohomological generalizations.

Original authors: Lie Fu, Zhiyuan Li, Teppei Takamatsu, Haitao Zou

Published 2026-04-13

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Lie Fu, Zhiyuan Li, Teppei Takamatsu, Haitao Zou

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 a detective trying to solve a mystery about a very exclusive, high-end neighborhood. This neighborhood is made up of complex geometric shapes called Hyper-Kähler varieties. These are like the "super-versions" of K3 surfaces (which are already famous, beautiful, 2D shapes in mathematics).

Your job is to answer a question posed by a famous mathematician named Shafarevich: "If we look at all the houses in this neighborhood that are built over a specific set of countries (number fields) and have 'good condition' (good reduction) everywhere except for a few specific bad roads (places), are there only a finite number of unique house designs?"

In simpler terms: If you limit where these shapes can live and how "broken" they can be, is the number of different shapes actually infinite, or is it finite?

Here is the breakdown of what the authors (Lie Fu, Zhiyuan Li, Teppei Takamatsu, and Haitao Zou) discovered, using some everyday analogies.

1. The Problem: Too Many "Twins" and "Renovations"

For a long time, mathematicians knew the answer was "Yes, it's finite" if the houses came with a specific, strict blueprint (called a polarization). Think of this as a house that must have a specific type of roof and a specific garden size. If you fix the roof and garden, there are only so many ways to build the house.

However, the authors wanted to solve the harder version: The Unpolarized Case.

The Analogy: Imagine you don't care about the roof or the garden. You just care about the structure of the house.
The Twist: In the world of Hyper-Kähler varieties, two houses can look completely different on the outside but be "birationally equivalent." This means you can take House A, knock down a few walls, and rearrange the rooms to get House B. They are essentially the same "soul" but different "bodies."
The Challenge: Because you can rearrange the walls (birational transformations) and because the "owners" (automorphisms) of these houses can be tricky, it seemed like there might be an infinite number of ways to twist and turn these shapes, making the set infinite.

2. The Solution: The "Uniform Kuga–Satake" Map

The authors' secret weapon is a tool they call the Uniform Kuga–Satake Map.

The Metaphor: Imagine you have a collection of very complex, abstract sculptures (the Hyper-Kähler varieties). It's hard to count them or tell them apart.
The Tool: The Kuga–Satake map is like a universal translator or a 3D scanner. It takes these complex sculptures and converts them into a different, more familiar object: Abelian Varieties (which are like multi-dimensional donuts or tori).
Why it works: We already know the rules for these "donuts." We know that if you limit the "donuts" to a specific area and condition, there are only finitely many of them.
The Innovation: Previous versions of this scanner only worked if the sculpture had a specific "polarization" (a specific tag). The authors built a Uniform scanner that works for any Hyper-Kähler variety, regardless of its tags. They proved that if the original sculpture has "good reduction" (is in good shape), the resulting "donut" also has good reduction.

3. The "Essentially Good" Compromise

There was a snag. Because these shapes can be rearranged (birational transformations), a shape might look "broken" (bad reduction) in one form, but if you rearrange the walls, it looks "perfect" (good reduction) in another form.

The Fix: The authors introduced a concept called "Essentially Good Reduction."
The Analogy: Imagine a house that looks like a wreck on the outside. But, if you hire a contractor to do some minor renovations (a finite, unramified extension), the house turns out to be structurally sound. The authors say, "If it can be fixed with a simple renovation, we count it as 'good'."
The Result: They proved that even with this flexibility, the number of unique "renovated" shapes is still finite.

4. The "Ghost" Problem (Cohomology)

The authors also tackled a deeper question: What if we only look at the "energy" or "vibrations" of the house (its cohomology) rather than the house itself?

The Trap: Sometimes, the "owners" of the house (the automorphism group) can hide. They can twist the house in a way that changes the house's identity but leaves its "vibrations" (cohomology) exactly the same. This creates an infinite number of "ghost houses" that look identical from the outside but are different inside.
The Solution: The authors showed that if the "owners" are honest (they act faithfully on the vibrations), the set is finite. If the owners are "ghosts" (non-faithful), the set can be infinite. However, they found a way to fix this by listening to more than just the main vibration (looking at higher-degree cohomology), which catches the ghosts.

5. The "Special Guests" (CM Type)

Finally, they looked at a special group of these shapes called CM type (Complex Multiplication). These are like the "VIPs" of the neighborhood—they have extra symmetry.

The Result: They proved that even if you let these VIPs live in any country with a limited population (bounded degree number field), there are only finitely many unique VIP designs. This also solved a side mystery about the "Brauer group" (a measure of the house's hidden secrets), proving these secrets are also limited in number.

Summary

In plain English, this paper says:

"Even though these complex geometric shapes can be rearranged and twisted in confusing ways, if you limit where they live and how 'broken' they are allowed to be, there are only a finite number of unique shapes. We proved this by building a new, universal machine that translates these complex shapes into simpler, well-understood 'donuts,' allowing us to count them easily."

This is a massive step forward in understanding the fundamental structure of these high-dimensional mathematical objects, unifying previous results and solving problems that were thought to be out of reach.

1. Problem Statement and Background

The Shafarevich conjecture (or question) asks whether the set of isomorphism classes of varieties in a specific family, defined over a number field $K$ and having good reduction outside a finite set of places $S$ , is finite.

Historical Context:
- Polarized Case: Faltings proved this for curves ( $g \ge 2$ ) and abelian varieties. André extended this to K3 surfaces and hyper-Kähler varieties, but only under the assumption of a polarization (an ample line bundle) with bounded degree.
- Unpolarized Case: The "unpolarized" version removes the restriction on the polarization. Y. She recently proved the unpolarized Shafarevich conjecture for K3 surfaces (dimension 2).
The Gap: For higher-dimensional hyper-Kähler varieties (dimension $2n \ge 4$ ), the unpolarized conjecture remained open.
Specific Challenges for Hyper-Kähler Varieties:
1. Birational Complexity: Unlike K3 surfaces, higher-dimensional hyper-Kähler varieties admit non-trivial birational transformations that are not isomorphisms. These transformations preserve the second cohomology but change the variety's isomorphism class.
2. Automorphism Groups: The action of the automorphism group on the second cohomology is not always faithful. This leads to "twists" that preserve cohomological properties (like unramifiedness) but create infinitely many distinct isomorphism classes, potentially violating finiteness.
3. Deformation Types: The finiteness of deformation classes in a fixed dimension is not fully known, necessitating a restriction to a fixed deformation type (e.g., $K3^{[n]}$ , $Kum_n$ , $OG_6$ , $OG_{10}$ ).

2. Methodology and Key Tools

The authors unify and generalize She's approach for K3 surfaces to higher dimensions using several sophisticated arithmetic and geometric tools:

A. Uniform Kuga–Satake Map

Concept: The classical Kuga–Satake construction associates a polarized hyper-Kähler variety to an abelian variety. However, the standard construction depends on the polarization degree.
Innovation: The authors construct a Uniform Kuga–Satake map inspired by She's work and Zarhin's trick. This map is independent of the polarization degree.
Mechanism: They embed the orthogonal lattice of the hyper-Kähler variety (specifically the orthogonal complement of the polarization) into a fixed unimodular lattice $\Sigma$ . This allows mapping the moduli space of hyper-Kähler varieties to a Shimura variety of orthogonal type, and subsequently to a Shimura variety of abelian type (Siegel modular variety) via the Kuga–Satake morphism.
Arithmetic Property: Crucially, they prove that if a hyper-Kähler variety has essentially good reduction (defined below), the associated Kuga–Satake abelian variety has good reduction. This allows the application of Faltings' finiteness theorems for abelian varieties.

B. "Essentially Good Reduction"

Definition: A hyper-Kähler variety $X$ has essentially good reduction at a place $v$ if, after a finite unramified extension of the base field, $X$ becomes birational to a variety with good reduction.
Motivation: Since birational hyper-Kähler varieties share the same second cohomology, the standard "good reduction" condition is too strict for the unpolarized case. The "essentially good reduction" condition accommodates birational modifications while preserving the necessary arithmetic constraints.

C. Cohomological Shafarevich Sets

The authors investigate sets where the condition is relaxed from "good reduction" to "unramifiedness of the second étale cohomology" ( $H^2_{\text{ét}}$ ).
They identify that if the automorphism group acts non-faithfully on $H^2$ , the naive cohomological conjecture fails (due to infinite twists).
Remedy: They propose a full-degree cohomological Shafarevich conjecture, requiring unramifiedness of cohomology in degrees other than 2 as well, which forces the automorphism group to be trivial (or finite) and restores finiteness.

D. Matsusaka–Mumford Theorem for Algebraic Spaces

To handle the specialization of birational maps in the arithmetic setting, the authors extend the classical Matsusaka–Mumford theorem to algebraic spaces. This ensures that birational maps between generic fibers of hyper-Kähler varieties extend to isomorphisms on the special fibers (outside a closed subset), provided the special fiber is not ruled.

3. Key Contributions and Results

Main Theorem: Unpolarized Shafarevich Conjecture (Theorem 1.4)

The authors prove that for a fixed deformation type $M$ of hyper-Kähler varieties (with $b_2 \neq 3$ ), the set of $F$ -birational isomorphism classes of varieties with essentially good reduction outside a finite set of places is finite.

If $b_2 \ge 5$ , the set of $F$ -isomorphism classes is also finite (due to the finiteness of birational models within a deformation type).
This generalizes She's result for K3 surfaces to all known higher-dimensional hyper-Kähler types.

Cohomological Shafarevich Conjecture (Theorem 1.7 & 8.3)

Failure of Naive Version: They demonstrate that the conjecture fails if one only requires $H^2$ to be unramified and the automorphism group acts non-faithfully on $H^2$ (e.g., for generalized Kummer types and $OG_6$ types).
Success with Full Cohomology: By requiring unramifiedness of cohomology in a set of degrees $I$ (including 2) such that the kernel of the action on these groups is trivial, finiteness is recovered.
Result: For all known deformation types and in dimension 4, the full-degree cohomological Shafarevich conjecture holds.

Finiteness of CM Type Varieties (Theorem 1.10)

Generalizing a result by Orr and Skorobogatov for K3 surfaces, the authors prove that the set of geometric isomorphism classes of hyper-Kähler varieties of CM type (Complex Multiplication) defined over number fields of bounded degree is finite.

Method: They utilize the Uniform Kuga–Satake map to relate CM hyper-Kähler varieties to CM abelian varieties, then apply the finiteness of CM points on Shimura varieties.
Corollary: This implies the uniform boundedness of Brauer groups for hyper-Kähler varieties of CM type.

Geometric Picard Lattices (Conjecture 1.8 & Corollary 1.11)

They establish the finiteness of isometry classes of geometric Picard lattices for hyper-Kähler varieties of CM type, strengthening the Bombieri–Lang conjecture context for these moduli spaces.

4. Significance and Impact

Resolution of a Major Open Problem: The paper provides the first proof of the unpolarized Shafarevich conjecture for higher-dimensional hyper-Kähler varieties, a significant step beyond the K3 surface case.
Arithmetic Geometry of Higher Dimensions: It demonstrates that the arithmetic properties of hyper-Kähler varieties (good reduction, CM type) are tightly controlled by their cohomological structures, even in the absence of a fixed polarization.
New Tools: The construction of the Uniform Kuga–Satake map is a major technical achievement. It decouples the arithmetic study from the specific polarization degree, allowing for a unified treatment of the moduli space.
Handling Birationality: The work rigorously addresses the issue of birational equivalence in higher dimensions, defining "essentially good reduction" as the correct arithmetic invariant for unpolarized families.
Broader Applications: The results on CM types and Brauer groups have implications for the arithmetic of moduli spaces and the distribution of rational points on hyper-Kähler varieties.

In summary, this paper successfully extends the powerful arithmetic finiteness results known for curves and K3 surfaces to the complex world of higher-dimensional hyper-Kähler varieties, overcoming the obstacles of birational geometry and non-faithful automorphism actions through the innovative use of uniform Kuga–Satake constructions and refined cohomological conditions.