🔢 mathematics

A positive density of elliptic curves are diophantine stable in certain Galois extensions

The paper proves that for any cyclic $p$ -extension of $\mathbb{Q}$ with $p \in \{3, 5\}$ , there exists an effectively computable positive density of elliptic curves defined over $\mathbb{Q}$ that remain Diophantine stable within that extension.

Original authors: Anwesh Ray, Pratiksha Shingavekar

Published 2026-06-03

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Anwesh Ray, Pratiksha Shingavekar

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 the world of numbers as a vast, infinite library. Inside this library, there are special books called Elliptic Curves. These aren't stories with characters; they are complex mathematical shapes defined by equations. Mathematicians love to study them because they hold secrets about prime numbers and the fundamental structure of arithmetic.

Now, imagine you have a specific "room" in this library, which we'll call a Number Field Extension (let's call it $L$ ). This room is built on top of the main library floor (the rational numbers, $\mathbb{Q}$ ).

The Big Question: Do New Points Appear?

Usually, when you move from the main floor to this special room $L$ , you might find new "points" (solutions to the equation) that didn't exist on the main floor. It's like walking from your living room into a new room and suddenly finding a chair that wasn't there before.

Diophantine Stability is a fancy way of saying: "No new chairs appeared."
If an elliptic curve is "Diophantine Stable" in room $L$ , it means every solution you can find in the big room $L$ was already sitting on the main floor $\mathbb{Q}$ . Nothing new was added.

The Paper's Discovery

For a long time, mathematicians asked: "If I pick one specific elliptic curve, can I find a room where it stays stable?" The answer was yes.

But in this paper, authors Anwesh Ray and Pratiksha Shingavekar flipped the question around. Instead of picking one curve and looking for a room, they said:

"Let's fix the room (a specific type of extension called a cyclic extension for primes 3 or 5). How many elliptic curves in the entire library are stable in this room?"

The Result: They proved that a positive density of these curves are stable.
In plain English: If you randomly pick an elliptic curve from the library (ordered by how "tall" or complex it is), there is a guaranteed, non-zero chance that it will be stable in this specific room. It's not just a few rare exceptions; it's a whole "neighborhood" of curves that behave this way.

How Did They Prove It? (The Detective Work)

To prove this, the authors used a multi-step strategy, which they describe using tools from algebra and statistics.

1. The "No Torsion" Filter (Checking for Ghosts)
First, they had to ensure the curves didn't have "torsion" points (points that loop back on themselves) in the new room. They used a powerful theorem (Duke's Theorem) which basically says: "Almost all elliptic curves are so complex that they don't have these looping points in small extensions." So, for 99.9% of curves, this condition is already met.

2. The "Rank" Filter (Checking for New Dimensions)
The harder part was proving the "rank" is zero. The rank is like the number of independent directions you can travel in the solution space. If the rank is 0, the solution space is tiny (just the points you started with).
To prove the rank stays 0, they looked at something called the Selmer Group. Think of the Selmer Group as a "security checkpoint" or a "filter."

If the filter is empty (size 0), it guarantees the rank is 0.
The authors showed that if a curve passes a few specific local tests (checking its behavior at specific prime numbers like 2, 3, or 5), then the filter will be empty in the new room.

3. The "Sieve" (Counting the Winners)
They constructed a giant "sieve" (a mathematical net) with specific holes.

They defined a set of rules (congruence conditions) that an elliptic curve must follow to pass the security checkpoint.
They used advanced counting techniques (from Bhargava and Shankar) to calculate how many curves pass through this sieve.
They found that the "density" of curves passing the sieve is a specific, positive number.

The "Recipe" for Stability

The paper gives a specific recipe for these stable curves. For a curve to be stable in this specific room $L$ :

It must behave nicely at the prime number 2 (which splits completely in the room).
It must behave nicely at the prime numbers that "ramify" (break apart) in the room.
It must satisfy a condition where the curve and its "twin" (a quadratic twist) have no points of order $p$ (where $p$ is 3 or 5) over the finite fields of those primes.

If a curve follows this recipe, the authors proved it is Diophantine Stable.

A Concrete Example

The paper even gives a real-world example to show the math works:

They chose the prime 3.
They chose a specific room $L$ related to the number 31 (the cubic subfield of the 31st roots of unity).
They calculated the exact "lower density" of curves that are stable in this room.
The result is a specific formula involving fractions like $1/4$ , $1/221$ , and a product of terms for every other prime number.

Summary

In simple terms, this paper is a census. It says: "If you build a specific type of mathematical extension (a room) using the number 3 or 5, there is a measurable, positive percentage of all elliptic curves that will not gain any new solutions when you enter that room. They remain exactly as they were."

The authors didn't just say "it happens sometimes"; they gave a precise mathematical formula for how often it happens, proving that these stable curves are a significant, permanent feature of the number system.

Technical Summary: A Positive Density of Elliptic Curves is Diophantine Stable in Certain Galois Extensions

Problem Statement
The paper addresses the problem of diophantine stability for elliptic curves. Following the definition by Mazur and Rubin [MR18], an elliptic curve $E$ defined over $\mathbb{Q}$ is diophantine stable in a number field extension $L/\mathbb{Q}$ if the natural inclusion induces an equality $E(L) = E(\mathbb{Q})$ . While previous work (e.g., [MR18]) established that for a fixed elliptic curve, there exist infinitely many cyclic extensions where the curve is diophantine stable, this paper reverses the perspective. The authors fix a specific cyclic $p$ -extension $L/\mathbb{Q}$ (where $p \in \{3, 5\}$ ) and investigate the density of elliptic curves $E/\mathbb{Q}$ , ordered by height, that are diophantine stable in this fixed extension. Specifically, they seek to prove the existence of a positive density of such curves where $E(L) = E(\mathbb{Q}) = 0$ .

Methodology
The authors employ a combination of Galois cohomology, arithmetic statistics, and sieve theory. The methodology proceeds in three main stages:

Cohomological Stability of Selmer Groups:
The authors establish conditions under which the $p$ -Selmer group of an elliptic curve remains trivial upon base change from $\mathbb{Q}$ to $L$ . Building on arguments by Česnavičius [Č17] and Brau [Bra14], they analyze the restriction map between Selmer groups $S_p(E/\mathbb{Q}) \to S_p(E/L)^G$ . They prove that if $S_p(E/\mathbb{Q}) = 0$ and specific local conditions are met at primes dividing $p$ and primes ramified in $L$ , then $S_p(E/L) = 0$ . This implies that the rank of $E(L)$ is 0 and the $p$ -torsion is trivial.
Arithmetic Statistics and Density Formulation:
To count the curves, the authors utilize the framework of Bhargava and Shankar [BS15a, BS15b, BS13], which relates the average size of $p$ -Selmer groups to the geometry of numbers. They define a "large family" of elliptic curves $F$ defined by local congruence conditions (congruence conditions on the coefficients $A$ and $B$ of the Weierstrass equation).
- They control the torsion subgroup by showing that for a density 1 set of curves, $E(L)_{\text{tors}} = 0$ , relying on Serre's Open Image Theorem and Duke's result [Duk97] regarding exceptional primes.
- They control the rank by constructing a family where the root number $\omega(E)$ is $+1$ for exactly 50% of the curves. By Theorem 3.11 (derived from Bhargava-Shankar), this ensures a positive proportion of curves in the family have $S_p(E/\mathbb{Q}) = 0$ .
Sieve Theoretic Arguments:
The authors define a specific family $\mathcal{E}$ of elliptic curves satisfying local conditions at primes ramifying in $L$ (denoted $\mathcal{Z}$ ) and at primes 2 and $p$ . These conditions ensure:
- Good reduction at $p$ and all $\ell \in \mathcal{Z}$ .
- Trivial $p$ -torsion for both $E$ and its quadratic twist $E^{-1}$ over the residue fields $\mathbb{F}_\ell$ for $\ell \in \mathcal{Z}$ .
- Specific reduction types at 2 (additive reduction) to control root numbers.
  Using Lemma 4.6 and explicit calculations (verified via Sagemath code in Remark 4.7), they demonstrate that the local conditions at ramified primes are satisfiable for sufficiently large primes (specifically $\ell \ge 5779$ for $p=5$ and $\ell \ge 233$ for $p=3$ ). They then apply a sieve argument (Proposition 4.8) to show that the intersection of these local conditions yields a family with a computable positive lower density.

Key Contributions and Results
The primary result is Theorem 3.8, which states:
Let $p \in \{3, 5\}$ and $L/\mathbb{Q}$ be a Galois extension with $\text{Gal}(L/\mathbb{Q}) \simeq \mathbb{Z}/p\mathbb{Z}$ . Let $\mathcal{Z}$ be the set of primes ramifying in $L$ . If:

$p \notin \mathcal{Z}$ and 2 splits completely in $L$ ;
For all $\ell \in \mathcal{Z}$ , there exists an elliptic curve over $\mathbb{F}_\ell$ with trivial $p$ -torsion for both the curve and its quadratic twist by $-1$;

Then there exists a positive density of elliptic curves $E/\mathbb{Q}$ such that $E(L) = E(\mathbb{Q}) = 0$ .

The paper provides an explicit lower bound for this density, given by the product:
$\eta_p \prod_{\ell} \delta_\ell$
where $\eta_p$ is $1/4$ for $p=3$ and $3/8$ for $p=5$ , and $\delta_\ell$ are local factors depending on the reduction type and ramification. For instance, in the case where $p=3$ and $L$ is the cubic subfield of $\mathbb{Q}(\mu_{31})$ , the authors compute a specific lower bound involving the term $585/961$ derived from the count of valid curves over $\mathbb{F}_{31}$ .

Significance and Claims
The authors frame this work as a "new direction of enquiry" in the study of diophantine stability. While previous literature focused on varying the field extension for a fixed curve, this paper fixes the extension and varies the curve. The significance lies in:

Quantitative Stability: Moving beyond existence proofs to establishing that diophantine stability is not a rare phenomenon but occurs with positive density in specific Galois extensions.
Explicit Bounds: Providing an effective, computable lower bound for this density, which depends on local invariants of the extension $L$ .
Methodological Synthesis: Successfully integrating the cohomological criteria for rank stability (Proposition 2.7) with the statistical machinery of Bhargava and Shankar to count curves satisfying these criteria.

The paper notes that the case $p=2$ is omitted as their methods do not apply, and the condition regarding the existence of curves with trivial $p$ -torsion over $\mathbb{F}_\ell$ (Condition 2) is verified computationally for large primes and theoretically for specific cases, ensuring the density is strictly positive.