Beilinson's conjecture on K3 surfaces with an involution

This paper proves that the Beilinson conjecture holds for specific K3 surfaces over $\bar{\mathbb{Q}}$ equipped with an involution, provided their quotient by the involution is a projective plane branched along a sextic curve.

The Gist

Here is an explanation of Kalyan Banerjee's paper, translated from the dense language of algebraic geometry into a story about maps, mirrors, and puzzles.

The Big Picture: What is this paper about?

Imagine you are trying to solve a massive, cosmic puzzle. In the world of mathematics, specifically Algebraic Geometry, there is a famous question called Beilinson's Conjecture.

Think of a K3 surface (the subject of this paper) as a complex, multi-dimensional shape made of pure math. It's like a very fancy, abstract donut, but with more holes and twists than you can count. Mathematicians want to know: How many different ways can we arrange "points" on this shape?

The paper proves that for a specific type of these shapes, the answer is surprisingly simple: There is effectively only one way to arrange them. The "mess" of points collapses into nothingness.

The Cast of Characters

To understand the proof, let's meet the main players using everyday analogies:

1. The K3 Surface ($S$): Imagine a complex, 3D sculpture made of clay. It's smooth and perfect.
2. The Involution ($i$): This is a magic mirror placed right in the middle of the sculpture. When you look at the sculpture in this mirror, it reflects itself perfectly. Some parts of the sculpture touch the mirror and stay still; other parts are swapped with their reflection.
3. The Quotient ($S/i$): If you take the sculpture and glue every point to its reflection in the mirror, you get a new, simpler shape. In this paper, that new shape is a flat sheet of paper (the Projective Plane, $\mathbb{P}^2$).
4. The Branch Locus (The Sextic): This is the "seam" where the mirror touches the sculpture. On the flat sheet of paper, this seam looks like a specific, complex flower pattern drawn with a six-petaled curve (a sextic).
5. Rational Curves: Think of these as straight lines or simple loops drawn on the complex sculpture. The paper assumes there are infinitely many of these simple lines hidden inside the complex shape.

The Problem: The "Infinite" Mess

Mathematicians have a tool called the Chow Group. Imagine this as a giant filing cabinet where you store every possible arrangement of points on your sculpture.

• For some shapes, this cabinet is tiny and organized.
• For others (like the K3 surfaces usually), the cabinet is infinite and chaotic. You can create infinite variations of point arrangements that are all mathematically distinct.

Beilinson's Conjecture predicts that if your shape is defined over "rational numbers" (like fractions, rather than infinite decimals), this cabinet should actually be empty or very small. It shouldn't be chaotic.

The Solution: The "Mirror Trick"

Banerjee's paper proves this conjecture for our specific K3 surface by using a clever "pincer movement" involving the mirror (the involution).

Step 1: The Mirror Says "Stay Put"
The author uses a technique developed by the famous mathematician Claire Voisin. The logic goes like this:

• The mirror ($i$) reflects the shape.
• Because the shape is a K3 surface, the mirror acts like a "do nothing" button on the most complex parts of the point arrangements.
• Analogy: Imagine you have a chaotic pile of sand. You put a mirror over it. The reflection of the sand is identical to the sand. The math says that if you try to move the sand using the mirror's rules, the sand doesn't actually move. The mirror acts as the Identity (it leaves things alone).

Step 2: The Flat Sheet Says "Flip It"
Now, look at the simpler shape we got by gluing the reflections together (the flat sheet of paper, $\mathbb{P}^2$).

• On a flat sheet of paper, the mirror acts differently. It acts like a negative sign.
• Analogy: If you write a number on a piece of paper and look at it in a mirror, the mirror flips it. In math terms, the mirror acts as -1.

Step 3: The Contradiction that Solves Everything
Here is the magic moment:

• The mirror tells the points on the K3 surface: "You are the same as your reflection" (Identity = $+1$).
• But the geometry of the surface also tells the points: "You are the opposite of your reflection" (because the quotient is a plane) (Identity = $-1$).

So, for any point arrangement $x$, we have:
$$x = -x$$

If a number is equal to its own negative, what is it? Zero.

The "Secret Sauce": Why this paper is special

You might ask, "Why didn't someone prove this 50 years ago?"

The paper works over $\bar{\mathbb{Q}}$ (the field of algebraic numbers). This is a very specific, countable universe of numbers.

• The Old Way (Over Complex Numbers): If you work with all real numbers (like $\mathbb{C}$), the "infinite lines" (rational curves) on the surface behave differently. The proof breaks down.
• The New Way (Over $\bar{\mathbb{Q}}$): Banerjee had to invent a new way to count and organize these infinite lines using Faltings' Theorem (a famous result about how points on curves behave).

He essentially showed that because there are infinitely many simple lines (rational curves) on the surface, and because we are working with "rational" numbers, the chaotic "infinite" possibilities collapse. The "infinite" lines force the "infinite" mess of points to shrink down until nothing is left but zero.

The Conclusion

The Main Result:
For these specific K3 surfaces (the ones that look like a double-covered flat sheet with a flower pattern), the "Chow Group" (the filing cabinet of point arrangements) is empty.

In Simple Terms:
If you take this specific type of mathematical shape, defined by rational numbers, and you try to shuffle its points around, you will find that every possible shuffle is actually the same as doing nothing. The shape is so rigid and symmetric that it has no "hidden complexity" left to discover.

Why it matters:
This confirms a major prediction (Beilinson's Conjecture) for a class of shapes that was previously unsolved. It shows that even in the most complex geometric worlds, if you have enough symmetry (the mirror) and the right kind of numbers (rational), the chaos disappears, leaving only perfect order.

Technical Summary

Here is a detailed technical summary of the paper "Beilinson's Conjecture on K3 Surfaces with an Involution" by Kalyan Banerjee.

1. Problem Statement

The paper addresses a specific instance of Beilinson's Conjecture regarding the Chow groups of zero-cycles on smooth projective surfaces defined over the field of algebraic numbers, $\overline{\mathbb{Q}}$.

• Context: For a smooth projective surface $S$ over $\mathbb{C}$, Mumford's theorem states that if the geometric genus $p_g > 0$, the Chow group of zero-cycles $CH_0(S)$ is infinite-dimensional. Conversely, Bloch's conjecture posits that if $p_g = 0$, $CH_0(S)$ is isomorphic to the points of an abelian variety (specifically, the Albanese variety).
• Beilinson's Conjecture: Over $\overline{\mathbb{Q}}$, the conjecture is stronger: for any smooth projective surface, the Albanese kernel is trivial. That is, the group of zero-cycles of degree zero modulo rational equivalence, denoted $A_0(S)$, is isomorphic to the Albanese variety. If the irregularity $q=0$ (as is the case for K3 surfaces), the conjecture implies $A_0(S) = 0$.
• Specific Target: The author investigates K3 surfaces $S$ defined over $\overline{\mathbb{Q}}$ equipped with an involution $i$ such that the quotient $S/i$ is isomorphic to the projective plane $\mathbb{P}^2$, and the covering map is branched along a sextic curve. The goal is to prove that $A_0(S) = 0$ for these surfaces, provided they contain infinitely many rational curves.

2. Methodology

The proof relies on a combination of algebraic geometry techniques adapted from complex geometry (specifically Voisin's work) and arithmetic geometry (Faltings' theorems). The methodology proceeds in three main stages:

A. Adaptation of Voisin's Technique to $\overline{\mathbb{Q}}$

Voisin previously proved that for complex K3 surfaces with an involution acting as the identity on the holomorphic 2-form, the involution acts as the identity on $A_0(S)$.

• Challenge: The standard techniques rely on properties of uncountable fields (like $\mathbb{C}$). Banerjee adapts these techniques to the countable algebraically closed field $\overline{\mathbb{Q}}$.
• Key Tool: The concept of finite dimensionality in the sense of Roitman. A subgroup $P \subset CH_0(X)$ is finite-dimensional if it is contained in the image of a correspondence from a smooth projective variety $W$.

B. The "Zero Map" Lemma (Theorem 2.2)

The core of the argument is a lemma stating that if a correspondence $Z$ on a surface $S$ (with $q=0$ and infinitely many rational curves) has an image in $CH_0(S)$ that is finite-dimensional (Roitman sense), then the induced map $Z^*: CH_0(S)_{hom} \to CH_0(S)$ is the zero map.

• Proof Strategy:
  1. Assume the image is contained in $\Gamma^*(W)$.
  2. Construct a curve $C \subset S$ (a hyperplane section) such that its Jacobian $J(C)$ is simple and has dimension larger than $\dim(W)$.
  3. Utilize the existence of infinitely many rational curves on $S$. This allows the construction of zero-cycles on $S$ derived from intersections of $C$ with these rational curves.
  4. Apply Faltings' Theorem (Mordell-Lang conjecture for abelian varieties): If the kernel of the map $J(C) \to A_0(S)$ were "small," it would contradict the density of torsion points or the finiteness of rational points on curves of genus $>1$.
  5. The argument concludes that the kernel must be large enough to force the map to be zero.

C. Analysis of the Involution

The author defines the correspondence $\Gamma = \Delta_S - \text{Graph}(i)$, where $\Delta_S$ is the diagonal and $i$ is the involution.

• Action on $A_0(S)$:
  1. Since the quotient is $\mathbb{P}^2$ (which has trivial $A_0$), the involution $i$ acts as $-1$ on $A_0(S)$.
  2. Using the adapted Voisin technique, the author proves that the image of the correspondence $\Gamma^*$ is finite-dimensional in the Roitman sense.
  3. By Theorem 2.2, since the image is finite-dimensional, the map induced by $\Gamma$ must be zero.
  4. Therefore, $(1 - i^*) = 0$ on $A_0(S)$, implying $i^*$ acts as the identity.

3. Key Contributions

1. Arithmetic Adaptation of Voisin's Theorem: The paper successfully extends Voisin's result on symplectic involutions acting trivially on $CH_0$ from the complex setting to the arithmetic setting ($\overline{\mathbb{Q}}$). This requires a novel proof of the "finite dimensionality implies zero map" lemma using Faltings' Theorem rather than complex analytic methods.
2. Explicit Class of K3 Surfaces: The paper identifies a concrete class of K3 surfaces (double covers of $\mathbb{P}^2$ branched along a sextic) where Beilinson's conjecture holds.
3. Role of Rational Curves: The proof explicitly utilizes the condition that the surface contains infinitely many rational curves. This condition is crucial for generating the necessary zero-cycles to apply Faltings' Theorem and force the kernel of the Albanese map to be trivial.

4. Main Results

Theorem 1.1 / Theorem 2.8:
Let $S$ be a K3 surface defined over $\overline{\mathbb{Q}}$ with an involution $i$ such that:

• The quotient $S/i \cong \mathbb{P}^2$.
• The branch locus is a sextic curve in $\mathbb{P}^2$.
• $S$ contains infinitely many rational curves.

Conclusion: The group $A_0(S)$ is trivial (i.e., $A_0(S) = \{0\}$). Consequently, Beilinson's conjecture holds for these surfaces.

Example 2.9: The author cites examples from [EJ] (Elsenhans & Jahnel) of K3 surfaces with Picard rank 1 and degree 2, which satisfy the conditions of the theorem, thereby providing concrete instances where the conjecture is verified.

5. Significance and Limitations

• Significance: This work bridges the gap between the geometric understanding of Chow groups (over $\mathbb{C}$) and the arithmetic understanding (over $\overline{\mathbb{Q}}$). It provides a rare verification of Beilinson's conjecture for K3 surfaces, which are generally difficult to handle due to their $p_g=1$ (though the Albanese kernel is the focus here).
• Limitation (Remark 2.10): The author explicitly warns that the result does not hold over $\mathbb{C}$ in the same manner. The proof relies intrinsically on Faltings' Theorem (finiteness of rational points on curves over number fields), which has no direct analogue over $\mathbb{C}$. Over $\mathbb{C}$, the group $A_0(S)$ for K3 surfaces is generally not known to be trivial, and the "finite dimensionality" argument used here fails because the field is uncountable.

Summary

Kalyan Banerjee proves that for a specific family of K3 surfaces over $\overline{\mathbb{Q}}$ (double covers of $\mathbb{P}^2$ branched along a sextic), the group of zero-cycles of degree zero is trivial. The proof hinges on showing that the involution acts as both $+1$ and $-1$ on this group, forcing it to be zero. This is achieved by adapting Voisin's geometric techniques to the arithmetic setting, utilizing the existence of infinitely many rational curves and Faltings' Theorem to demonstrate that specific correspondences induce the zero map on the Chow group.