Alexander Valov,Baruch Meerson,Pavel V. Sasorov
Abstract:We evaluate, in the large-$N$ limit, the complete probability distribution $\mathcal{P}(A,m)$ of the values $A$ of the sum $\sum_{i=1}^{N} |\lambda_i|^m$, where $\lambda_i$ ($i=1,2,\dots, N$) are the eigenvalues of a Gaussian random matrix, and $m$ is a positive real number. Combining the Coulomb gas method with numerical simulations using a matrix variant of the Wang-Landau algorithm, we found that, in the limit of $N\to \infty$, the rate function of $\mathcal{P}(A,m)$ exhibits phase transitions of different characters. The phase diagram of the system on the $(A,m)$ plane is surprisingly rich, as it includes three regions: (i) a region with a single-interval support of the optimal spectrum of eigenvalues, (ii) a region emerging for $m<2$ where the optimal spectrum splits into two separate intervals, and (iii) a region emerging for $m>2$ where the maximum or minimum eigenvalue ``evaporates" from the rest of eigenvalues and dominates the statistics of $A$. The phase transition between regions (i) and (iii) is of second order. Analytical arguments and numerical simulations strongly suggest that the phase transition between regions (i) and (ii) is of (in general) fractional order $p=1+1/|m-1|$, where $0<m<2$. The transition becomes of infinite order in the special case of $m=1$, where we provide a more complete analytical and numerical description. Remarkably, the transition between regions (i) and (ii) for $m\leq 1$ and the transition between regions (i) and (iii) for $m>2$ occur at the ground state of the Coulomb gas which corresponds to the Wigner's semicircular distribution.
What problem does this paper attempt to address?
### The problems the paper attempts to solve
This paper aims to study the probability distribution problem of the linear statistics of eigenvalues of Gaussian random matrices under large deviations. Specifically, the author focuses on the linear statistics in the following form:
\[
A = \sum_{i = 1}^N|\lambda_i|^m
\]
where $\lambda_i$ are the eigenvalues of an $N\times N$ Gaussian random matrix, and $m$ is a positive real number. The author's goal is to calculate the complete probability distribution $P(A, m)$ of the linear statistic $A$ in the limit as $N\to\infty$.
### Main problems and methods
1. **Large - deviation theory**: The author uses large - deviation theory to study the tail behavior of the linear statistic $A$, that is, the probability distribution when $A$ is far from its expected value.
2. **Coulomb gas method**: To solve this problem, the author uses the Coulomb gas method (Coulomb gas method), interpreting the joint probability distribution of eigenvalues as the energy distribution of a two - dimensional Coulomb gas.
3. **Numerical simulation**: Combined with numerical simulation, especially using the matrix - variant Wang - Landau algorithm, the theoretical predictions are verified, and the phase - transition behavior under different parameters is further explored.
### Key findings
1. **Phase diagram**: On the $(A, m)$ plane, the author discovers a rich phase diagram, including three regions:
- Region (i): Optimal spectral eigenvalues with single - interval support.
- Region (ii): For $m < 2$, the optimal spectrum splits into two independent intervals.
- Region (iii): For $m > 2$, the largest or smallest eigenvalue "evaporates" out and dominates the statistical properties of $A$.
2. **Phase - transition types**:
- The phase transition between region (i) and region (iii) is a second - order phase transition.
- The phase transition between region (i) and region (ii) is a fractional - order phase transition with an order of $p = 1+\frac{1}{|m - 1|}$. In particular, when $m = 1$, the phase transition is of infinite order.
3. **Critical behavior**: For $0 < m\leq1$, the phase transition occurs at the ground state of the Coulomb gas, that is, $A=\bar{A}$ corresponding to the Wigner semicircle distribution.
4. **Single - particle scaling**: For $m > 2$, when $A > \bar{A}$, the probability distribution $P(A, m)$ exhibits single - particle scaling behavior rather than collective scaling behavior. This change is caused by the "evaporation" phenomenon of eigenvalues, where one eigenvalue separates from the rest of the eigenvalues and dominates the linear statistic $A$.
### Summary
By combining analytical methods and numerical simulations, this paper studies in detail the large - deviation behavior and phase - transition characteristics of the linear statistics of eigenvalues of Gaussian random matrices, revealing the rich phase diagrams and critical behaviors of the system under different parameters. These results not only deepen the understanding of random matrix theory but also provide a new perspective for the study of other strongly correlated systems.