7
$\begingroup$

Assume that $n \geq 3$. Is there a subgroup $H \leq {\rm GL}_{2}(\mathbb{Z}/2^{n} \mathbb{Z})$ whose order is a power of $2$ so that

$\bullet$ $\det : H \to (\mathbb{Z}/2^{n} \mathbb{Z})^{\times}$ is surjective, and

$\bullet$ $H/\Phi(H) \cong (\mathbb{Z}/2\mathbb{Z})^{2}$, and

$\bullet$ $H$ contains $-I = \begin{bmatrix} -1 & 0 \\ 0 & -1 \end{bmatrix}$, and

$\bullet$ $H$ contains all matrices of the form $I + 2^{n-1} X$ for $X \in M_{2}(\mathbb{F}_{2})$?

Motivation: Suppose that $E$ is an elliptic curve defined over a number field $K$, $E(K)$ contains a point of order $2$, and $K \cap \mathbb{Q}(\zeta_{8}) = \mathbb{Q}$. What is the smallest number of quadratic subextensions of $K(E[2^{n}])/K$? The smallest possible is $3$, namely $K(i)$, $K(\sqrt{2})$ and $K(\sqrt{-2})$. The elliptic curves for which this occurs have the property that for any positive integer $n$, the image of the mod $2^{n}$ Galois representation is a subgroup of ${\rm GL}_{2}(\mathbb{Z}/2^{n} \mathbb{Z})$ that satisfies the first and second conditions above. If $E/K$ is non-CM, there is a positive integer $n$ for which this subgroup contains all matrices $\equiv I \pmod{2^{n-1}}$. For more detailed information and context, see the paper here.

Is there ever a case where $E$ and all of its quadratic twists have this property? This would occur if $-I$ is in the image of the mod $2^{n}$ Galois representation attached to $E$, which is the motivation for the fourth condition.

Note: I have found all subgroups satisfying the first three conditions for $3 \leq n \leq 9$ and none satisfy the fourth condition. I noticed that for $5 \leq n \leq 9$, if $H \leq {\rm GL}_{2}(\mathbb{Z}/2^{n} \mathbb{Z})$ satisfies the first three conditions, then the maximal subgroup $$ \{ g \in H : \det(g) \equiv \pm 1 \pmod{8} \} $$ is abelian, which forces $H$ to be rather small.

$\endgroup$
5
  • $\begingroup$ What its $\Phi$ ? Are you sure you want $X\in M_2 ({\mathbb F}_2 )$ ? $\endgroup$ Commented Feb 21, 2024 at 17:54
  • $\begingroup$ Here $\Phi(H)$ is the Frattini subgroup of $H$, the intersection of all maximal subgroups of $H$. Also, if $X$ is defined mod $2$, then the entries of $2^{n-1} X$ make sense mod $2^{n}$. Another way of wording condition 4 is to say that $H$ contains all matrices $\equiv I \pmod{2^{n-1}}$. $\endgroup$ Commented Feb 21, 2024 at 18:00
  • $\begingroup$ OK, thanks you ! $\endgroup$ Commented Feb 21, 2024 at 19:04
  • $\begingroup$ If I’m not mistaken, the first two conditions imply that $\Phi(H)=\{h \in H,\, \det{h} \equiv 1 \pmod{8}\}$, or, to put it another way, that no proper subgroup of $H$ has surjective determinant. I’m however not sure how it could help. $\endgroup$ Commented Feb 21, 2024 at 21:16
  • $\begingroup$ Yes, you are correct that the given conditions mean no proper subgroup has surjective determinant. $\endgroup$ Commented Feb 21, 2024 at 21:28

0

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.