The generalized prime number theorem for automorphic \(L\)-functions (Q1044812): Difference between revisions

Let \(\pi \) and \(\tilde {\pi }\) be automorphic irreducible cuspidal representations of GL\(_m (\mathbb{Q}_{\mathbb{A}})\) and \(L(s, \pi \times \tilde {\pi })\) be the Rankin-Selberg \(L\)-function. For complete definitions see the paper. Consider \[ \frac{L^{(k)}(s, \pi \times \tilde {\pi })}{L(s, \pi \times \tilde {\pi })}=(-1)^{k}\sum_{n=1}^{\infty}\frac{\rho_{\pi \times \tilde {\pi }}(n)}{n^s}, \quad \sigma>1, \] where \(k\) is a positive integer. Assume that \(\pi\) is self-contragredient: \(\pi \cong \tilde {\pi }\). The main result of the paper states that \[ \sum_{n\leq x}\rho_{\pi \times \tilde {\pi }}(n)=(k\log^{k-1}x+a_{1,k}\log^{k-2}x+\dots+a_{k-1,k})x+O\left(x\exp(-c\sqrt{\log x})\right), \] where the complex constants \(a_{j,k}\) (\(j=1,\dots,k-1\)) are computable. This result generalizes the corresponding result (for \(k=1\)) of \textit{J. Liu} and \textit{Y. Ye} [Pure Appl. Math. Q. 3, No. 2, 481--497 (2007; Zbl 1246.11152)].