Algebraic cycles on Gushel-Mukai varieties

Basic Information

• Paper ID: 2207.01118
• Title: Algebraic cycles on Gushel-Mukai varieties
• Authors: Lie Fu (Université de Strasbourg), Ben Moonen (Radboud University Nijmegen)
• Classification: math.AG (Algebraic Geometry)
• Publication Date: Épijournal de Géométrie Algébrique, Special volume in honour of C. Voisin, Article No. 17 (2024)
• Paper Link: https://arxiv.org/abs/2207.01118

Abstract

This paper investigates algebraic cycles on Gushel-Mukai (GM) varieties over the complex numbers. The authors prove the generalized Hodge conjecture, the (motivated) Mumford-Tate conjecture, and the generalized Tate conjecture for all GM varieties. They compute all integral Chow groups of GM varieties, except for two exceptional infinite-dimensional cases (1-cycles on GM fourfolds and 2-cycles on GM sixfolds). The paper establishes that if two GM varieties are generalized partners or generalized duals, then their rational Chow motives of intermediate degree are isomorphic.

Research Background and Motivation

Problem Background

1. Importance of Gushel-Mukai varieties: GM varieties constitute a class of Fano varieties with rich geometric structure, deeply connected to hyperkähler varieties. They can be explicitly defined as intersections of the cone CGr(2,V₅) with linear subspaces and quadratic forms.
2. Core problems in algebraic cycle theory: This paper addresses fundamental questions about algebraic cycles on GM varieties, including:
   • Hodge conjecture: the relationship between algebraic cycle classes and Hodge classes
   • Tate conjecture: the relationship between algebraic cycles and Galois actions
   • Mumford-Tate conjecture: a bridge connecting Hodge theory and Galois theory
3. Research significance: These conjectures are central problems in algebraic geometry, crucial for understanding the geometric and arithmetic properties of algebraic varieties.
4. Existing limitations: Although these conjectures are known for certain special types of varieties, a complete study for GM varieties had not been accomplished previously.

Core Contributions

1. Proof of three major conjectures for GM varieties:
   • Generalized Hodge conjecture for all GM varieties
   • Mumford-Tate conjecture for even-dimensional GM varieties
   • Generalized Tate conjecture for even-dimensional GM varieties
2. Complete computation of Chow groups: Calculation of integral Chow groups for all GM varieties, except for two infinite-dimensional cases
3. Establishment of motivic isomorphisms: Proof that GM varieties that are generalized partners or generalized duals have isomorphic rational Chow motives of intermediate degree
4. Foundation for positive characteristic results: These results serve as an important foundation for the authors' proof of the Tate conjecture for GM varieties in characteristic p≥5 in the accompanying paper FM22

Detailed Methods

Basic Theory of GM Varieties

Definition: An n-dimensional GM variety (n∈{3,4,5,6}) is a smooth projective variety realizable as the intersection:

X = CGr(2,V₅) ∩ P(W) ∩ Q

where CGr(2,V₅)⊂P(ℂ⊕∧²V₅) is the cone over the Grassmannian, W⊂ℂ⊕∧²V₅ is a linear subspace of dimension n+5, and Q is a quadric.

Classification: GM varieties fall into two types:

• Mukai type: When O∉P(W), the Gushel map γ is a closed embedding
• Gushel type: When O∈P(W), γ is a double cover onto its image

Core Techniques for Chow Group Computation

1. Application of Bloch-Srinivas Theorem

For GM threefolds and certain higher-dimensional cases, the Bloch-Srinivas theorem is applied:

• When CH₀(X) is supported on an r-dimensional subset with r≤2, then Griff²(X)=0
• This implies that algebraic equivalence coincides with homological equivalence

2. Innovative Method for GM Sixfolds

Rational chain connectedness of the variety of lines:

• Proof that the variety of lines F₁(X) on a GM sixfold X is rationally chain connected
• Using the technique of successive ruled surfaces, every 1-cycle is shown to be equivalent to an integer multiple of a line class

Technical details:

Key Lemma: Any two points on X can be connected by at most 4 lines
Proof strategy: Utilizes geometric properties of the Gushel bundle UX and 
relative Hilbert schemes

3. Proof of Homological Triviality

For 3-cycles on GM sixfolds, the coniveau spectral sequence and unramified cohomology theory are employed:

• Application of the Gersten conjecture of Bloch-Ogus
• Use of results by Colliot-Thélène and Voisin based on the Bloch-Kato conjecture

Unified Framework for Conjecture Proofs

Hodge Conjecture

Strategy: Primarily relies on existing literature results and Chow group computations

• GM dimensions 3, 4, 5: References work of Laterveer and others
• GM dimension 6: Follows directly from the Chow group computations in this paper

Mumford-Tate Conjecture

André's criterion: Verification of conditions A_i and B_i^+:

• Condition A_i: Intermediate cohomology is of K3 type ✓
• Condition B_i^+: Existence of a connected family such that the image of the period map contains an open subset of the period domain ✓

Generalized Tate Conjecture

Logical chain:

Generalized Hodge Conjecture + Mumford-Tate Conjecture → Generalized Tate Conjecture

Experimental Setup

Theoretical Verification Framework

This is a pure theoretical research paper, with "experiments" manifested in:

1. Concrete computational verification: Detailed Chow group calculations for GM varieties of different dimensions
2. Case analysis: Analysis of geometric properties of special GM varieties
3. Consistency checking: Verification that results obtained by different methods are consistent

Computational Tools

• Cohomology theories: Hodge theory, crystalline cohomology
• Algebraic K-theory: Application of Quillen's higher K-groups
• Derived categories: Utilization of equivalences of Kuznetsov components

Experimental Results

Main Theorems

Theorem 4.5 (Chow groups of GM sixfolds):

• CH₀(X) ≅ CH₁(X) ≅ CH₆(X) ≅ ℤ
• CH₅(X) → H₁₀(X,ℤ)(5) is an isomorphism
• CH₂(X) → H₄(X,ℤ)(2) is an isomorphism
• CH₃(X) → H₆(X,ℤ)(3) is injective

Theorem 6.2 (Proof of conjectures): Let X be an even-dimensional complex GM variety. Then:

• The Mumford-Tate conjecture holds
• The generalized Tate conjecture holds

Theorem 8.2 (Motivic isomorphisms): If X and X' are generalized partners or generalized duals of GM varieties, then:

h^n(X) ≅ h^n'(X')((n'-n)/2)

Specific Computational Results

GM Threefolds (Theorem 4.1)

• CH^i(X) ≅ ℤ, i=0,1,3
• Griff²(X) = 0
• 0 → J(ℂ) → CH²(X) → ℤ → 0 (split exact sequence)

GM Fivefolds (Theorem 4.4, result of Zhou)

• CH₀(X) ≅ CH₁(X) ≅ CH₄(X) ≅ CH₅(X) ≅ ℤ
• CH₂(X) ≅ ℤ⊕ℤ
• Griff³(X) = 0

Related Work

Historical Development

1. Bloch-Srinivas (1983): Established fundamental connections between Chow groups and geometric properties
2. Laterveer's series of works: Proved related conjectures for certain cases of GM varieties
3. Debarre-Kuznetsov series: Established systematic theory of GM varieties

Innovations in This Paper

1. Completeness: First comprehensive picture of algebraic cycle theory for GM varieties
2. Technical breakthroughs: Innovative methods in the GM sixfold case
3. Unified framework: Incorporates all three major conjectures into a unified proof framework

Conclusions and Discussion

Main Conclusions

1. Theoretical completeness: Essentially resolves core problems concerning algebraic cycles on GM varieties
2. Method universality: Developed techniques can be generalized to other Fano varieties
3. Applied value: Provides foundation for research in positive characteristic

Limitations

1. Dimension restriction: Only considers cases 3≤dim≤6
2. Characteristic restriction: Main results limited to characteristic 0
3. Computational complexity: Two infinite-dimensional Chow groups remain incompletely determined

Future Directions

1. Positive characteristic generalization: Partially realized in accompanying paper FM22
2. Higher-dimensional cases: Possible generalizations of GM varieties
3. Computational methods: New techniques for handling infinite-dimensional Chow groups

In-Depth Evaluation

Strengths

1. Theoretical depth: Unified resolution of multiple profound conjectures, demonstrating exceptional technical mastery
2. Methodological innovation: Particularly the "successive ruled surfaces" technique in the GM sixfold case
3. Completeness: Provides a complete picture of algebraic cycle theory for GM varieties
4. Applied value: Provides important foundation for subsequent research, particularly in positive characteristic

Weaknesses

1. Technical threshold: Requires deep background in algebraic geometry; limited accessibility
2. Computational complexity: Certain proof procedures are highly technical and difficult to simplify
3. Unresolved problems: Two infinite-dimensional cases remain unresolved

Impact

1. Academic contribution: Major progress on core problems in algebraic geometry
2. Methodological influence: Developed techniques applicable to other geometric objects
3. Subsequent research: Has already catalyzed breakthrough work in positive characteristic

Applicable Scenarios

1. Theoretical research: Research in algebraic geometry, motivic theory, and Hodge theory
2. Technical application: Provides methodological guidance for studying other Fano varieties
3. Pedagogical value: Important case study in algebraic cycle theory

References

This paper cites important literature in algebraic geometry, including:

• Foundational work of Bloch-Srinivas on algebraic cycles
• Systematic study of GM varieties by Debarre-Kuznetsov
• André's criterion for the Mumford-Tate conjecture
• Deep work on Hodge theory by Voisin and others

This paper represents an important milestone in the study of GM varieties in algebraic geometry. It not only resolves multiple core conjectures but also provides important technical and theoretical foundations for the development of related fields.