Toward Scalable Normalizing Flows for the Hubbard Model
This work addresses scalability issues in generative modeling for condensed matter physics, but it appears incremental as it builds on existing methods without introducing major new paradigms.
The authors tackled the challenge of scaling normalizing flows for the Hubbard model to larger lattices and lower temperatures, focusing on stability and efficiency improvements, and presented scaling behaviors for stochastic normalizing flows and non-equilibrium MCMC methods.
Normalizing flows have recently demonstrated the ability to learn the Boltzmann distribution of the Hubbard model, opening new avenues for generative modeling in condensed matter physics. In this work, we investigate the steps required to extend such simulations to larger lattice sizes and lower temperatures, with a focus on enhancing stability and efficiency. Additionally, we present the scaling behavior of stochastic normalizing flows and non-equilibrium Markov chain Monte Carlo methods for this fermionic system.