arXiv: Data Structures and Algorithms: On zeros and algorithms for disordered systems: mean-field spin glasses
Authors: Ferenc Bencs, Kuikui Liu, Guus Regts
Spin glasses are fundamental probability distributions at the core of statistical physics, the theory of average-case computational complexity, and modern high-dimensional statistical inference. In the mean-field setting, we design deterministic quasipolynomial-time algorithms for estimating the partition function to arbitrarily high accuracy for nearly all inverse temperatures in the second moment regime. In particular, for the Sherrington–Kirkpatrick model, our algorithms succeed for almost the entire replica-symmetric phase. To achieve this, we study the locations of the zeros of the partition function. Notably, our methods are conceptually simple, and apply equally well to the spherical case and the case of Ising spins.