Towards Keating-Snaith’s conjecture for cubic Hecke L L -functions over the Eisenstein field
Hua Lin, Peng-Jie Wong · 2025
Read Towards Keating-Snaith’s conjecture for cubic Hecke L L -functions over the Eisenstein field by Hua Lin, Peng-Jie Wong (2025) 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-2511-08783.txt.
Introduction
Hua Lin and Peng-Jie Wong work toward the Keating-Snaith conjecture, a famous prediction about the statistical distribution of L-function values at their central point. The conjecture says these values follow a log-normal law whose mean and variance are fixed by the symmetry type of the family.
The family here is thin and specific: cubic Hecke L-functions over the Eisenstein field. Building on recent work of Radziwill and Soundararajan, the authors prove a conditional lower bound in the direction of the conjecture. A key new input estimates the one-level density of low zeros for these functions in a twisted setting, extending earlier work by David and Güloğlu. The result assumes the Generalized Riemann Hypothesis, so it stands as progress toward the conjecture rather than a settlement of it.
Catalog record
| CREATOR | Hua Lin, Peng-Jie Wong |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 83 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2511-08783
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Towards Keating-Snaith’s conjecture for cubic Hecke L L -functions over the Eisenstein field Thanks: PJW is currently supported by the NSTC grant 114-2628-M-110-006-MY4. Hua Lin, Peng-Jie Wong Abstract . A famous conjecture of Keating and Snaith asserts that central values of L L -functions in a given family admit a log-normal distribution with a prescribed mean and variance depending on the symmetry type of the family. Based on a recent work of Radziwiłł and Soundararajan, we obtain a conditional lower bound towards Keating-Snaith’s conjecture for a “thin” family of cubic Hecke L L -functions over the Eisenstein field. A key new input is certain twisted estimates of the 1-level density of zeros of cubic Hecke L L -functions, extending the previous work of David and Güloğlu, under the Generalised Riemann Hypothesis. 2000 Mathematics Subject Classification Primary 11M06; Secondary 11M41 1. Introduction About a century ago, Hilbert and Pólya independently suggested a connection between zeros of the Riemann zeta function and eigenvalues of some self-adjoint operator. There was little evidence at the time, but since the encounter between Montgomery and Dyson about half a century ago, the Riemann zeta function and L -functions have been intimately linked to random matrix theory. Using this connection, Keating and Snaith conjectured that central values of L -functions in families admit a log-normal distribution with a prescribed mean and variance depending on their symmetry type [ 16 ] . Their far-reaching conjecture presents a natural analogue of Selberg’s central limit theorem on the log-normality of the Riemann zeta function on the critical line {\mathfrak{Re}}(s)=\frac{1}{2} and lies in the centre of the theory of L -functions. For instance, Keating-Snaith’s conjecture implies that almost all the central values of L -functions in families are non-vanishing, which is unknown in general. In the same vein, prior to the work of Keating and Snaith, Katz and Sarnak [ 13 , 14 ] analysed random matrices and monodromy for families of L -functions over function fields. Their work not only shed light on determining the symmetry type of families of L -functions, but also led to their famous density conjecture on the distribution of “low-lying” zeros of L -functions. Their conjecture, known as Katz and Sarnak’s philosophy, implies the non-vanishing of families of L -functions at the central point. For example, some progress has been made for families of L -functions of modular forms by Iwaniec, Luo, and Sarnak [ 12 ] , and for families of L -functions of quadratic twists of elliptic curves by Heath-Brown [ 9 ] . Most families that have been studied, however, are over the rationals \mathbb{Q} and are “self-dual”. Nonetheless, Cho and Park studied the low-lying zeros of cubic Dirichlet L -functions in [ 2 ] and recently, David and Güloğlu [ 4 ] studied the 1-level density of zeros for cubic Hecke L -functions over the Eisenstein field \mathcal{K}=\mathbb{Q}(\omega) , where \omega=e^{2\pi i/3} , under the Generalised Riemann Hypothesis (GRH). For the “thin” family \mathcal{F} defined in ( 2.1 ), using the work of Heath-Brown [ 8 ] , David and Güloğlu obtained a support for the Fourier transform of the test function beyond the (-1,1) -barrier and achieved at least \frac{2}{13} non-vanishing at the central point, supporting Katz and Sarnak’s philosophy. It is also worth mentioning that by calculating the moments of cubic Hecke L -functions, it has been shown in [ 1 , 6 ] that there are infinitely many non-vanishing central values. Moreover, positive propositions of non-vanishing at the central point in the same family were recently established in [ 7 ] (under GRH) and [ 3 ] (unconditionally). We refer the reader to these references for more details. More recently, Radziwiłł and Soundararajan [ 18 ] proved a conditional lower bound towards Keating-Snaith’s conjecture for quadratic twists of elliptic curves with global root number 1. This was extended by the second-named author to quadratic twists with global root number -1 in [ 20 ] . The innovation of Radziwiłł and Soundararajan is a method of obtaining lower bounds on the non-vanishing of central L -values via studying “twisted” 1-level density estimates for low-lying zeros and using the “method of moments” from probability theory. As an illustration, they established conditional lower bounds towards Keating-Snaith’s conjecture by “upgrading” the work of Heath-Brown [ 9 ] . In this article, we extend the method of Radziwiłł and Soundararajan [ 18 ] to the unitary family \mathcal{F} of cubic Hecke characters over \mathbb{Z}[\omega] , considered by David and Güloğlu [ 4 ] (see also ( 2.1 )), and prove the following theorem towards Keating-Snaith’s conjecture (cf. [ 15 , pg. 59, (iii)] ). Theorem 1.1 . Assume the Generalised Riemann Hypothesis (GRH) for L(s,\chi_{\mathfrak{f}}) with \mathfrak{f}\in\mathcal{F} . For any fixed (\alpha,\beta) , we have as X\rightarrow\infty , \displaystyle\begin{split}\#\bigg\{\mathfrak{f}\in\mathcal{F}&:X<\mathrm{N}(\mathfrak{f})\leq 2X,\frac{\log|L(\frac{1}{2},\chi_{\mathfrak{f}})|}{\sqrt{\log\log\mathrm{N}(\mathfrak{f})}}\in(\alpha,\beta)\bigg\}\\ &\geq\frac{2}{13}\left(\int_{\alpha}^{\beta}\frac{1}{\sqrt{2\pi}}e^{-\frac{1}{2}t^{2}}dt+o(1)\right)\#\{\mathfrak{f}\in\mathcal{F}:X<\mathrm{N}(\mathfrak{f})\leq 2X\},\end{split} where the integrand is the probability density function of a standard normal random variable. Remark . (i) This is the first result of this type for a family of non-rational and non-self-dual L -functions to the best of the authors’ knowledge. (ii) Compared to the argument of David and Güloğlu [ 4 ] , our analysis of the twisted 1-level density requires a more delicate treatment of twisted cubic Gauss sums where conductors are no longer prime powers. As seen in Section 5.3 , this is the core of our argument. 2. Notation and preliminaries Let \omega=e^{2\pi i/3}=\frac{-1+\sqrt{-3}}{2} be a primitive third root of unity, and let \mathcal{K}=\mathbb{Q}(\omega) . The ring of integer \mathcal{O}_{\mathcal{K}}=\mathbb{Z}[\omega] , where every element n can be expressed as n=a+b\omega for some a,b\in\mathbb{Z}. Since \mathbb{Z}[\omega] is a principal ideal domain, every n coprime to 3 has a unique generator n=(m) , m\equiv 1(\operatorname{mod}3). We use this fact freely throughout the article. Also, we let \Lambda_{\mathcal{K}},\phi_{\mathcal{K}}, and \mu_{\mathcal{K}} denote the number field analogues of the corresponding arithmetic functions (namely, \Lambda(n),\phi(n), and \mu(n) ) over \mathcal{K}. We denote by \mathrm{N}(n) the norm of n\in\mathbb{Z}[\omega] , and \mathrm{N}(n)=n\bar{n} , where \bar{n} is the complex conjugate of n. Furthermore, if n=a+b\omega, then one has \displaystyle\mathrm{N}(n)=(a+b\omega)(a+b\bar{\omega})=a^{2}+ab+b^{2}\in\mathbb{Z}. The cubic Dirichlet character \chi_{p} modulo a prime p\equiv 1(\operatorname{mod}3)\in\mathbb{Z}[\omega] is the cubic residue symbol \displaystyle\chi_{p}(n)=\left(\frac{n}{p}\right)_{3}\equiv n^{\frac{\mathrm{N}(p)-1}{3}}(\operatorname{mod}p), and cubic characters of general moduli n are defined multiplicatively. We recall the following statement for cubic reciprocity. Lemma 2.1 . Let m,n\in\mathbb{Z}[\omega] be coprime to 3 . If m,n\equiv\pm 1(\operatorname{mod}3) , then \displaystyle\left(\frac{m}{n}\right)_{3}=\left(\frac{n}{m}\right)_{3}. A character is primitive if it is the product of \chi_{p} or \overline{\chi_{p}}=\chi_{p^{2}} for distinct primes p . A cubic Dirichlet character is a cubic Hecke character if \chi_{n}(\omega)=1. This is true if and only if \mathrm{N}(n)\equiv 1(\operatorname{mod}9) , so primitive Hecke characters have conductors n=f_{1}f_{2}^{2}, where f_{1},f_{2} are square-free, with (f_{1},f_{2})=1 and \mathrm{N}(f_{1}f_{2}^{2})\equiv 1(\operatorname{mod}9). The thin family \mathcal{F} of cubic Hecke characters over \mathbb{Z}[\omega] is, as considered by David and Güloğlu in [ 4 ] , (2.1) \displaystyle\mathcal{F}=\{\mathfrak{f}\in\mathbb{Z}[\omega]:\mathfrak{f}\neq 1,\mathfrak{f}\equiv 1(\operatorname{mod}9),\mathfrak{f}\text{ square-free}\}. Let \chi_{\mathfrak{f}} be a primitive Hecke character with \mathfrak{f}\neq 1 and (\mathfrak{f},3)=1. The completed L -function is \displaystyle\Lambda(s,\chi_{\mathfrak{f}})=(|\Delta_{\mathcal{K}}|\mathrm{N}(\mathfrak{f}))^{s/2}(2\pi)^{-s}\Gamma(s)L(s,\chi_{\mathfrak{f}}), where \Delta_{\mathcal{K}}=-3 is the discriminant of \mathcal{K}=\mathbb{Q}(\omega) . By the work of Hecke, it is known that \Lambda(s,\chi_{\mathfrak{f}}) is entire and it satisfies the functional equation \displaystyle\Lambda(s,\chi_{\mathfrak{f}})=\chi_{\mathfrak{f}}(\sqrt{-3})g(1,\mathfrak{f})\mathrm{N}(\mathfrak{f})^{-1/2}\Lambda(1-s,\overline{\chi_{\mathfrak{f}}}), where g(1,\mathfrak{f}) is the generalized cubic Gauss sum defined in ( 2.4 ). Let h be a smooth