Title: Critical probability distributions of the order parameter from the functional renormalization group

Abstract: We show that the functional renormalization group (FRG) allows for the calculation of the probability distribution function of the sum of strongly correlated random variables. On the example of the three-dimensional Ising model at criticality and using the simplest implementation of the FRG, we compute the probability distribution functions of the order parameter or equivalently its logarithm, called the rate functions in large deviations theory. We compute the entire family of universal scaling functions, obtained in the limit where the system size $L$ and the correlation length of the infinite system $\xi_{\infty}$ diverge, with the ratio $\zeta=L/\xi_{\infty}$ held fixed. It compares very accurately with numerical simulations.