Simulations of light nuclei with neutral network wave functions

Neural-networks quantum states is now widely used by the physics community to solve few and many- body problems in various fields. The application to nuclear physics problems is however complicated by the nature of the interaction. In this talk we will present the a variational Monte Carlo method based on a new, highly-expressive, neural-network quantum state ansatz that permits to solve the few- and many-body nuclear Schroedinger equation in a systematical and improvable way. In particular we will focus on the computation of ground-state properties of atomic nuclei with up to A = 20 nucleons, using as input a leading order pionless effective field theory Hamiltonian. Finally we will present an innovative approach that permits to extract the electromagnetic properties as well as electroweak transitions from the neural-network quantum state wave functions.