Rational points on modular curves: parameterization and geometric explanations

Conditional on Zywina's effective Serre uniformity conjecture, this paper establishes a natural parameterization of non-CM rational points on all modular curves using finitely many such curves, thereby confirming Mazur and Ogg's philosophy that these points arise from the geometry of modular curves.

The Gist

Imagine you are a detective trying to solve a massive mystery: Where do all the "rational points" on a specific family of geometric shapes come from?

In the world of mathematics, these shapes are called Modular Curves. They are like a vast, infinite library of maps. Each map tells a story about a specific type of Elliptic Curve (a special kind of equation that looks like a twisted loop).

For decades, mathematicians have been trying to answer a question posed by the legendary Barry Mazur: "Can we classify every single point on every single one of these maps?"

This paper, by Derickx, Hashimoto, Najman, and Shnidman, is like a master key that finally unlocks the library, but with a few important caveats. Here is the breakdown in simple terms:

1. The Problem: A Library with Infinite Shelves

Think of the "Modular Curves" as an infinite library.

• The Books: Each book is a curve.
• The Pages: The "rational points" on the curve are specific locations on the map.
• The Mystery: Most of these books have pages that are empty or have only a few dots. But some books have infinite dots. The mathematicians wanted to know: Is there a pattern? Do these dots appear randomly, or is there a hidden rule?

For a long time, it seemed like the dots appeared randomly. But the authors realized there was a hidden structure, like a secret filing system.

2. The Solution: The "Twist" Family

The authors discovered that you don't need to look at every single book individually. Instead, you can group them into Families.

Imagine you have a master template (a "Base Curve").

• The Twist: You can take this template and "twist" it slightly (like changing the color of the paper or the font).
• The Result: Even though the twisted versions look different, they are all related to the original template.

The paper proves that almost all the rational points on these infinite curves come from just 160 specific "Base Curves."

• If you find a point on a weird, twisted curve, it's actually just a "twisted" version of a point you could have found on one of these 160 base curves.
• It's like realizing that every unique snowflake in a blizzard is just a slightly twisted version of one of 160 basic crystal shapes.

3. The "Oddballs": The 41 Isolated Points

While 160 families cover almost everything, the authors found 41 special "oddball" numbers (called $j$-invariants).

• These are like the "black sheep" of the family. They don't belong to any infinite group.
• They are isolated. You can't twist a curve to get them; they just exist on their own.
• The paper lists exactly these 41 numbers. If you find an elliptic curve with one of these numbers, you know it's a rare, isolated event that doesn't fit the general pattern.

4. The "Geometric Explanation": Why Do They Exist?

The most philosophical part of the paper answers a question from the mathematician Ogg: "Why do these points exist at all? Is there a geometric reason?"

The authors say: Yes.
They created a set of rules (like a game of "connect the dots") to explain every single point.

• Special Points: Some points are "special" (like the corners of the map). They are explained by their own nature.
• Push-Forward: If you have a point on a small map, and you project it onto a bigger map, the new point is explained by the old one.
• Collinearity: If three points lie on a straight line, and two are explained, the third one is explained too.
• Fiber Splicing: If two different maps intersect at exactly one point, that point is explained by the intersection.

The Big Conclusion:
They proved that every single rational point on every modular curve can be explained by these geometric rules. There are no "magic" points that appear out of nowhere. If a point exists, there is a geometric reason for it, just like if a tree grows, there is a reason (soil, sun, water).

5. The Catch: The "Conditional" Clause

There is one big "If" in this story.
The entire proof relies on a famous, unproven guess called Serre's Uniformity Conjecture.

• Think of this conjecture as a "Rule of the Universe" that mathematicians are 99% sure is true, but haven't proven yet.
• If this rule is true, then the authors' map is perfect.
• If the rule is false, then there might be some hidden "ghost" points we haven't found yet.

Summary Analogy

Imagine you are trying to catalog every possible shape a cloud can take.

• The Old Way: You try to draw every single cloud you've ever seen. Impossible.
• The New Way (This Paper): You realize that 99% of clouds are just variations of 160 basic shapes (twisted by wind). You also found 41 very rare, weird clouds that don't fit any pattern.
• The Explanation: You prove that every single cloud you see can be explained by the wind, the sun, and the atmosphere. There are no "ghost clouds" appearing out of thin air.
• The Caveat: This only works if we assume the laws of physics (Serre's Conjecture) are exactly as we think they are.

In a nutshell: The authors have organized the chaotic universe of these mathematical curves into a neat, understandable system, showing that almost everything comes from a few simple families, and everything that exists has a logical, geometric reason for being there.

Technical Summary

Here is a detailed technical summary of the paper "Rational Points on Modular Curves: Parameterization and Geometric Explanations" by Derickx, Hashimoto, Najman, and Shnidman.

1. Problem Statement

The paper addresses Mazur's "Program B," a broad program initiated in 1977 to classify the rational points on all modular curves $X_G$ over a number field $K$. Specifically, for $K=\mathbb{Q}$, the goal is to classify the images of adelic Galois representations $\rho_E: \text{Gal}_{\mathbb{Q}} \to \text{GL}_2(\widehat{\mathbb{Z}})$ for non-CM elliptic curves $E/\mathbb{Q}$.

The authors reformulate Program B into three precise questions:

1. Algorithmic: Can we describe the set $X_G(\mathbb{Q})$ given $G$?
2. Parameterization: Can we parameterize all $\mathbb{Q}$-rational points on all modular curves (or equivalently, all possible Galois images $G_E$) in terms of a finite set of data?
3. Geometric Characterization: Do all rational points on modular curves arise from "geometric" reasons (e.g., cusps, CM points, or specific geometric operations), confirming a philosophy by Mazur and Ogg?

The central challenge is that there are infinitely many modular curves of genus $g > 1$ with non-special rational points, making a direct classification seemingly impossible without a unifying structure.

2. Methodology

The authors build upon the recent work of David Zywina, who conjectured that the adelic images of elliptic curves over $\mathbb{Q}$ are constrained by a finite set of "agreeable" modular curves. The methodology involves:

• Twist Families and Agreeable Closures: The authors utilize the concept of the agreeable closure $G^{\text{ag}}$ of a subgroup $G \leq \text{GL}_2(\widehat{\mathbb{Z}})$. They define a map $\pi: X_G \to X_{G^{\text{ag}}}$ which is a torsor under a finite abelian group of "toric operators" $T_G$.
• Twist-Parameterization: They introduce the distinction between twist-parameterized and twist-isolated curves.
  • A curve $X_G$ is twist-parameterized if it is an abelian cover of a modular curve $X_H$ (where $H$ is agreeable) that has infinitely many rational points. In this case, the rational points on $X_G$ are parameterized by the rational points on $X_H$ and a character $\chi: \text{Gal}_{\mathbb{Q}} \to T_G$.
  • A curve is twist-isolated if it does not fit into such a family.
• Conditional Assumptions: The main results are conditional on Conjecture 4.7, which combines:
  • Serre's Uniformity Conjecture (Explicit Version): For $\ell > 13$, the mod $\ell$ image is surjective unless $j(E)$ belongs to a specific finite list.
  • Zywina's Conjecture: All non-special rational points on agreeable modular curves of finite rational points correspond to a specific finite set of "exceptional" $j$-invariants ($J_{\text{exc}}$).
• Geometric Axioms: To address the geometric characterization, the authors define a set of "explained points" generated by five axioms:
  1. Special points (cusps, CM).
  2. Push-forward under modular morphisms.
  3. Abelian lifts (pulling back explained points along abelian covers).
  4. Collinearity (using divisors on curves of genus 0, 1, and $>1$).
  5. Fiber splicing (intersection of fibers of maps to elliptic curves).

3. Key Contributions and Results

A. Parameterization of Galois Images (Theorem 2)

Assuming Conjecture 4.7, the authors prove that the adelic Galois images $G_E$ of all non-CM elliptic curves over $\mathbb{Q}$ are parameterized by the rational points on 160 explicit modular curves ($X_{M_i}$).

• Structure: For any non-CM $E$, there exists an index $i \in \{1, \dots, 160\}$ and a rational point $x \in X_{M_i}(\mathbb{Q})$ such that $G_E$ is conjugate to a twist $K_i^\chi$ of a specific subgroup $K_i$.
• Optimality: The set of 160 curves is minimal; no smaller set of modular curves can parameterize all possible Galois images.
• Classification:
  • 138 curves ($i=1 \dots 138$) have infinitely many rational points. These correspond to twist-parameterized families. The Galois images here vary in infinite families determined by the rational points on the base curve.
  • 22 curves ($i=139 \dots 160$) have finitely many rational points. These correspond to twist-isolated families.

B. Twist-Isolated $j$-invariants (Theorem 3)

The authors identify exactly 41 twist-isolated $j$-invariants (listed in Table 2).

• These correspond to the 22 modular curves with finite rational points.
• For these 41 $j$-invariants (up to quadratic twist), the associated Galois representations do not vary in an infinite algebraic family. They represent "sporadic" behavior in the landscape of elliptic curves.

C. Geometric Explanation of Rational Points (Theorem 4)

The paper provides a conditional proof that all rational points on all modular curves are "explained by geometry."

• Reduction: The problem reduces to explaining the rational points on the 22 twist-isolated curves (since points on the 138 infinite families are explained by the base curves and abelian lifts).
• Verification: The authors manually verify the geometric explanation for the 41 isolated $j$-invariants using the five axioms.
  • Example: For the genus 3 curve $X_{S_4(13)}$, points are explained via collinearity with CM points and fiber splicing on a genus 10 double cover.
• Significance: This confirms the Mazur-Ogg philosophy that rational points exist only when there is a geometric reason (e.g., they are generated by cusps, CM points, or specific geometric relations like collinearity).

D. Non-Serre Curves (Theorem 9.1)

The authors apply their classification to non-Serre curves (elliptic curves where the Galois image index is $>2$). They show these curves fall into 20 distinct families, each corresponding to a specific moduli interpretation (e.g., existence of an $\ell$-isogeny, lying in a normalizer of a Cartan subgroup, or specific field containment relations like $\mathbb{Q}(E[2]) \subset \mathbb{Q}(E[3])$).

E. Super-Sporadic Points (Section 10)

The paper constructs infinite families of super-sporadic points (points that remain sporadic on any modular cover). Using the twist families identified in Theorem 2, they show that for 95 of the 138 infinite families, rational points lift to super-sporadic points on higher-level curves.

4. Significance

• Resolution of Program B (Conditional): The paper provides the sharpest possible conditional solution to Mazur's Program B over $\mathbb{Q}$. It reduces the infinite complexity of all modular curves to a finite set of 160 curves and a finite set of 41 exceptional $j$-invariants.
• Bridging Algorithm and Theory: It demonstrates that while a general algorithm to find rational points on genus $g>1$ curves is an open problem, the specific structure of modular curves allows for a "parameterization" that bypasses the need for individual curve-by-curve analysis in most cases.
• Geometric Philosophy: The work rigorously validates the idea that "sporadic" rational points on modular curves are not random but are the result of specific geometric mechanisms (collinearity, fiber splicing, etc.).
• Computational Framework: The authors provide explicit tables (Tables 2 and 4) and code (available on GitHub) that allow researchers to determine the Galois image of any non-CM elliptic curve and check if a $j$-invariant is twist-isolated.

In summary, the paper transforms the classification of rational points on modular curves from an intractable infinite problem into a finite, structured classification, conditional on standard conjectures in number theory, and provides a deep geometric explanation for the existence of these points.