Boundary CM points and class groups of small exponent
David Aiken, Erick Ross, Dmitriy Shvydkoy, Hui Xue · 2026
Read Boundary CM points and class groups of small exponent by David Aiken, Erick Ross, Dmitriy Shvydkoy, Hui Xue (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-26020.txt.
Introduction
Aiken, Ross, Shvydkoy and Xue study complex multiplication points on the boundary of the classic fundamental domain for SL2(Z). Duke showed such points spread evenly across the domain as discriminants tend to minus infinity. The authors prove the boundary points also spread evenly along the boundary. They further pin down exactly when every CM point of a given discriminant lies on the boundary. Along the way, they classify, under an assumption, all negative discriminants whose class group has small exponent. The work links point distribution to algebraic structure.
Catalog record
| CREATOR | David Aiken, Erick Ross, Dmitriy Shvydkoy, Hui Xue |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 32584 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.26020
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Boundary CM points and class groups of small exponent David Aiken, Erick Ross, Dmitriy Shvydkoy, Hui Xue Abstract . Let ℱ \mathcal{F} denote the fundamental domain for SL 2 ( ℤ ) \text{SL}_{2}(\mathbb{Z}) on the upper half plane ℋ \mathcal{H} . William Duke showed that as fundamental discriminants D → − ∞ D\to-\infty , the sets CM D \textnormal{CM}_{D} (CM points of discriminant D D ) are equidistributed in ℱ \mathcal{F} . In this paper, we investigate the behavior of CM points on the boundary of ℱ \mathcal{F} . We prove that such CM points are equidistributed on the boundary, and also give a complete characterization of when every CM D \textnormal{CM}_{D} point lies on the boundary. Along the way, we also (conditionally) give a complete classification of negative discriminants with class group of small exponent. 1991 Mathematics Subject Classification 11G15, 11R65, 11R29 1. Introduction Let \mathcal{H}\coloneq\{z\in\mathbb{C}:\text{Im }z>0\} denote the complex upper half plane, and \mathcal{F}\coloneq\left\{z=x+iy\in\mathcal{H}:|z|\geq 1,-\frac{1}{2}\leq x\leq 0\right\}\cup\left\{z=x+iy\in\mathcal{H}:|z|>1,0<x<\frac{1}{2}\right\} denote the standard fundamental domain for \text{SL}_{2}(\mathbb{Z}) acting on \mathcal{H} . For negative discriminants D , let \textnormal{CM}_{D} denote the set of CM points in \mathcal{H} of discriminant D . It turns out that the action of \text{SL}_{2}(\mathbb{Z}) on \mathcal{H} is also an action on these \textnormal{CM}_{D} points. Hereafter, when we refer to “discriminants” in this paper, we are only referring to negative discriminants. In [ 3 ] , Duke investigated how CM points are distributed over the fundamental domain \mathcal{F} . In particular, he showed that for fundamental discriminants D ; as D\to-\infty , the \textnormal{CM}_{D} points are equidistributed via the hyperbolic measure d\mu_{\textnormal{hyp}}=\frac{1}{y^{2}}\,dx\,dy . In this paper, we investigate the behavior of CM points on the boundary of \mathcal{F} . We study this question for all discriminants D , not just the fundamental ones. First, we show that the CM points on the boundary of \mathcal{F} are equidistributed according to certain metrics. We will parametrize CM points on the left boundary of \mathcal{F} (i.e. \{-\frac{1}{2}+iy\ :\ y\geq\frac{\sqrt{3}}{2}\} ) by their imaginary parts y , and CM points on the lower arc of \mathcal{F} (i.e. \{e^{i\theta}\ :\ \frac{\pi}{2}\leq\theta\leq\frac{2\pi}{3}\} ) by their arguments \theta . Theorem 1.1 . Let \textnormal{CM}^{\textnormal{LB}}_{|D|\leq\Delta} denote the CM points on the left boundary of \mathcal{F} with absolute discriminant |D|\leq\Delta , and \textnormal{CM}^{\textnormal{LA}}_{|D|\leq\Delta} denote the CM points on the lower arc of \mathcal{F} with absolute discriminant |D|\leq\Delta . Then as \Delta\to\infty ; (1) \textnormal{Im\,}\textnormal{CM}^{\textnormal{LB}}_{|D|\leq\Delta} is equidistributed over every fixed [\frac{\sqrt{3}}{2},K] via the metric d\mu=\frac{1}{y}dy , (2) \textnormal{arg\,}\textnormal{CM}^{\textnormal{LA}}_{|D|\leq\Delta} is equidistributed over [\frac{\pi}{2},\frac{2\pi}{3}] via the metric d\nu=\frac{1}{\sin\theta}\,d\theta . We make two remarks here about this result. (1) This equidistribution theorem differs from Duke’s result in that we are considering \textnormal{CM}_{D} points for all discriminants |D|\leq\Delta , instead of for each individual discriminant D . This is necessary because for individual discriminants D , the \textnormal{CM}_{D} points do not even become dense along the boundary of \mathcal{F} (for example, there is only one \textnormal{CM}_{D} point on the boundary of \mathcal{F} for negative prime discriminants D ). (2) The two parts of Theorem 1.1 can actually be viewed as the same phenomenon. The two metrics d\mu and d\nu are precisely the restriction of the hyperbolic metric ds=\frac{\sqrt{dx^{2}+dy^{2}}}{y} to the left boundary of \mathcal{F} and to the lower arc of \mathcal{F} , respectively. And in fact, we will prove both parts of Theorem 1.1 in a unified manner (by transforming the lower arc of \mathcal{F} into the vertical line segment \{-\frac{1}{2}+iy:\frac{1}{2}\leq y\leq\frac{\sqrt{3}}{2}\} ). Next, in Theorem 1.2 , we give a classification of when all the \textnormal{CM}_{D} points in \mathcal{F} lie on the boundary of \mathcal{F} . This theorem utilizes the fact that the \textnormal{CM}_{D} points lying on the boundary of \mathcal{F} are precisely those of order dividing 2 (by Lemma 4.1 ). Theorem 1.2 . Every \textnormal{CM}_{D} point contained in \mathcal{F} lies on the boundary of \mathcal{F} if and only if \textnormal{Cl}(\mathcal{O}_{D}) has exponent dividing 2 and D is odd or -4 . Conditionally on the non-existence of Siegel zeros, such discriminants D are precisely those given in the following table. Unconditionally, this table could possibly also include the discriminants arising from one additional fundamental discriminant. Class Group Discriminants \{\textnormal{e}\} -3,-4,-7,-11,-19,-27,-43,-67,-163 \mathbb{Z}/2\mathbb{Z} -15,-35,-51,-75,-91,-99,-115,-123,-147,-187,-235,-267,-403,-427 (\mathbb{Z}/2\mathbb{Z})^{2} -195,-315,-435,-483,-555,-595,-627,-715,-795,-1435 (\mathbb{Z}/2\mathbb{Z})^{3} -1155,-1995,-3003,-3315 In order to show the second half of this theorem, we will need a classification of class groups \textnormal{Cl}(\mathcal{O}_{D}) of exponent dividing 2 . (Recall here that the exponent E of a group G is the minimal integer E\geq 1 such that g^{E}=e for all g\in G .) The methods we develop to give this classification turn out to also work for any given exponent E . So we give a (conditional) classification of all class groups \textnormal{Cl}(\mathcal{O}_{D}) with exponent 1\leq E\leq 8 . Such a classification was already given for fundamental discriminants in [ 4 ] . To extend from fundamental discriminants to all discriminants, we use the ideas from a recent work of Fan and Pollack [ 5 ] . In particular, for any given fundamental discriminant D_{0} , we use ideas from [ 5 ] to develop an algorithm to compute the full list of conductors f for D_{0} such that \textnormal{Cl}(\mathcal{O}_{D_{0}f^{2}}) has exponent E . Computing discriminants via this algorithm then yields Tables A.1 - A.8 , listing discriminants with class group of exponent 1\leq E\leq 8 . The following theorem gives conditions for which one can prove that these tables are complete. (And in the authors’ opinion, it is extremely likely that all of these tables are in fact complete.) Theorem 1.3 . Tables A.1 - A.8 list negative discriminants with class group of exponent 1\leq E\leq 8 . (1) Unconditionally, the table for exponent E=1 is complete. (2) Assuming the non-existence of Siegel zeros, the tables for exponent E=2,4,8 are all complete. (3) Unconditionally, the tables for exponent E=2,4,8 could also contain the discriminants arising from at most one additional fundamental discriminant. (4) Assuming ERH (Extended Riemann Hypothesis), the tables for exponent E=2,3,4,5,8 are all complete. (5) Assuming ERH, the table for E=6 is complete if there are no fundamental discriminants 3.1\cdot 10^{20}\leq|D|\leq 2.5\cdot 10^{25} with class groups of exponent 6 . (6) Assuming ERH, the table for E=7 is complete if there are no fundamental discriminants 3.1\cdot 10^{20}\leq|D|\leq 3.9\cdot 10^{30} with class groups of exponents 7 . Now, we would like to point out three different motivations for this project investigating CM points on the boundary of \mathcal{F} . First, this project is a natural variation of Duke’s celebrated result in [ 3 ] concerning the equidistribution of CM points over \mathcal{F} . Theorem 1.1 here gives an analogous result about the equidistribution of CM points over the boundary of \mathcal{F} . Second, it leads to a classification of all discriminants of small exponent, which is of independent interest (compare, for example, with [ 4 ] ). Third, the results of this paper were needed in [ 1 ] to show certain results about transcendence of zeros of modular forms. (In fact, this was the original reason why we started to investigate this topic of CM points on the boundary of \mathcal{F} .) It turns out that in certain cases, if a modular form has a non-transcendental zero, then it has zeros at every \textnormal{CM}_{D} point