Quantum Many-body Dynamics with Artificial Neural Networks

This lecture explores quantum many-body dynamics in two dimensions using artificial neural networks, focusing on experimental simulations and theoretical challenges like tensor network states. It discusses the use of convolutional neural networks as neural quantum states, regularization techniques, and the training process involving the Fubini-Study metric and the time-dependent variational principle. The results of quenching experiments in the transverse-field Ising model are presented, highlighting the out-of-equilibrium behavior and the impact of different initial conditions. The lecture also covers the ensembling approach, regularization methods, and the implications of noise in the S-matrix and F-vector. Key concepts include quantum phase transitions, entanglement growth, and the application of neural networks in quantum dynamics.