Markov Chain Monte Carlo: Detailed Balance & Neural Networks

This lecture introduces Markov chain Monte Carlo as a strategy to generate samples in a chain, explaining the transition probabilities and the detailed balance condition. It then delves into the application of neural networks in quantum states representation, discussing the normalization of wave functions and autoregressive quantum states. The instructor showcases the use of neural networks in approximating ground states of frustrated spins systems, highlighting the improvements in accuracy over time and the importance of symmetries in enhancing results.