Abstract:Let $\chi$ be a primitive Dirichlet character whose conductor $q$ is a prime number. For the certain averages of values of $\log |L(s, \chi)|$ in $q$-aspect at a fixed $s=\sigma>1/2$, under Generalized Riemann Hypothesis (GRH), we explain it can be written as integrals involving the same density function ($M$-function) for the average of values of the difference between the logarithms of two symmetric power $L$-functions in the level aspect.
For the distribution of values of $\log L(s, \chi_D)$ in the $D$-aspect at a fixed $s=\sigma>1/2$, where $\chi_D$ is a real character attached to a fundamental discriminant $D$, we construct a $M$-function under GRH. In the case $\sigma\geq 1$, we can obtain it unconditionally.
What problem does this paper attempt to address?
This paper attempts to solve the problems regarding the distribution of L - functions in the modulus aspect and the D - aspect. Specifically:
1. **Modulus Aspect (Modulus Aspect)**:
- For a real number \(\sigma>1/2\) and the logarithmic absolute value \(\log|L(\sigma,\chi)|\) of the primitive Dirichlet character \(\chi\), under the Generalized Riemann Hypothesis (GRH), the author shows that its average value can be represented by an integral, which involves the same density function (M - function) constructed in the level aspect of the logarithmic value difference of two symmetric - power L - functions. The specific formula is:
\[
\lim_{q\rightarrow\infty}\frac{1}{\phi(q)}\sum_{\chi\ mod\ q}^{\prime}\Phi(\log |L(\sigma,\chi)|)=\int_{-\infty}^{\infty}M_{\sigma}(u)\Phi(u)\,du\sqrt{2\pi}
\]
where \(\Phi\) is a bounded continuous function.
2. **D - Aspect (D - Aspect)**:
- For the logarithmic value \(\log L(\sigma,\chi_D)\) of the real character \(\chi_D\) attached to the fundamental discriminant \(D\), under the GRH, the author proves that there exists a density function \(Q_{\sigma}(x)\) such that:
\[
\lim_{Y\rightarrow\infty}\frac{1}{N(Y)}\sum_{|D|\leq Y}^{\ast}\Phi(\log L(\sigma,\chi_D))=\int_{-\infty}^{\infty}Q_{\sigma}(x)\Phi(x)\,dx\sqrt{2\pi}
\]
where \(N(Y)=\sum_{|D|\leq Y}^{\ast}1\). In particular, in the case of \(\sigma\geq1\), this result can be obtained without relying on the GRH.
### Main Contributions
- **Uniformity of M - functions**: The author proves that the M - functions in the modulus aspect and the D - aspect are the same by similar methods.
- **Unconditional Results**: For the case of \(\sigma\geq1\), the author is able to obtain the density function in the D - aspect without relying on the GRH.
### Methodology
- **Methods of Ihara and Matsumoto**: The author draws on the work of Ihara and Matsumoto, especially using similar strategies when dealing with M - functions.
- **Methods of Mourtada and Murty**: For the problems in the D - aspect, the author adopts the methods of Mourtada and Murty, using tools such as Dirichlet series expansion and Sato - Tate measure.
Through these methods, the author not only solves the distribution problems of L - functions in different aspects but also provides new perspectives and tools for further research.