Universality of correlation functions of Hermitian random matrices in an external field (Q1270197)

It is known that the eigenvalue distribution of random matrices \(M_N\) with the density of the probability distribution \[ Z_N^{-1} \exp\bigl\{- N \text{Tr} V(M_N) \bigr\} \] can be expressed in terms of the kernel \[ K_N (\lambda,\mu) =\sum^{N-1}_{j=0} F_j (\lambda) F_j (\mu), \] where \(\{F_j (\lambda)\}\) is the set of polynomials that are orthogonal with respect to the weight \(\exp\{-v(\lambda)\}\). The author derives a matrix integral representation for \(K_N (\lambda, \mu)\) that avoids use of the orthogonal polynomials. This allows him to investigate random matrix ensembles with the probability density \[ (Z_N')^{-1} \exp \bigl\{-N \text{Tr} V(M) +N \text{Tr} MA\bigr\} \] and to find universalities of the eigenvalue distribution in the limit \(N\to\infty\) and the short distance regime \(\lambda- \mu= O(N^{-1})\).