Nonzero coefficients of half-integral weight modular forms mod \(\ell \) (Q2319735): Difference between revisions

From MaRDI portal
Added link to MaRDI item.
RedirectionBot (talk | contribs)
Removed claim: author (P16): Item:Q449700
Property / author
 
Property / author: Kannan Soundararajan / rank
Normal rank
 

Revision as of 04:33, 15 February 2024

scientific article
Language Label Description Also known as
English
Nonzero coefficients of half-integral weight modular forms mod \(\ell \)
scientific article

    Statements

    Nonzero coefficients of half-integral weight modular forms mod \(\ell \) (English)
    0 references
    0 references
    0 references
    20 August 2019
    0 references
    Let \(K\) be a number field with ring of integers \(\mathcal{O}\). Let \(\ell\) be a prime number and \(\lambda\) a maximal ideal of \(\mathcal{O}\) above \(\ell\). Let \(f\) be a weakly holomorphic modular form of (non integral) half-integral weight and level \(\Gamma_1(N)\) for some integer \(N\geq 1\). Assume that the coefficients \(a_n\) of \(f\) lie all in \(\mathcal{O}\) and that, if \(\ell\geq 3\) then \(f\not\equiv 0\pmod{\lambda}\) whereas if \(\ell=2\), then \(f\) is not a constant modulo \(\lambda\). The authors prove the lower bound \[ \#\left\{n\leq X \colon a_n\not\equiv 0\pmod{\lambda}\right\}\gg C\frac{\sqrt{X}}{\log\log X}\qquad (X\to+\infty) \] for some \(C>0\) depending on \(f\). The expected result is the lower bound \(\sqrt{X}\), which is attained for a one-variable theta-series, and indeed, the lower bound \(X\) is expected for most forms of half-integral weight. This improves a result due to \textit{S. Ahlgren} and \textit{M. Boylan} [Math. Res. Lett. 15, No. 2-3, 409--418 (2008; Zbl 1234.11056)] who proved the lower bound with \((\log X)/(\log\log X)^A\) for any \(A\) instead of \(\log\log X\). The proof of the main result rests on the multiplication of \(f\) by a holomorphic cusp form of half-integral weight that has few non zero coefficients, to obtain a holomorphic cusp form of integral weight, and the proof of the following (optimal) result. Let \(u\) be an integer. For any integer \(X\) large enough, choose a subset \(\mathcal{A}_X\) of \(\{1,\dotsc, X\}\). Then \[ \begin{split} \#\{p\text{ prime }\colon pu\leq X, & \, \exists(a,m)\in\mathcal{A}_X\times\mathbb{N} \qquad pu=a+m^2\}\\ &\ll\frac{X^{1/2}}{\log X}\left(\#\mathcal{A}_X\log\log X+\#\mathcal{A}_X^{1/2}X^{1/4}\right) \end{split} \] as \(X\to+\infty\).
    0 references
    half-integral weight
    0 references
    modular form
    0 references
    coefficients
    0 references
    congruences
    0 references
    theta-function
    0 references
    partition function
    0 references
    proportion
    0 references
    density
    0 references

    Identifiers