Auxiliary-Field Methods for Quantum Materials - IPAM at UCLA

Explore auxiliary-field quantum Monte Carlo (AFQMC) methods for quantum materials in this 49-minute lecture presented by Shiwei Zhang from the Flatiron Institute's Center for Computational Quantum Physics. Gain insights into the connections between AFQMC and other quantum Monte Carlo methods, as well as its relation to neural networks. Discover recent algorithmic advances in ab initio simulations of solids, including correlated sampling, computation of gradients (forces and stresses), and structural optimization. Delve into topics such as the many-electron problem, challenges in accurate calculations, and strategies to understand and control the sign problem. Learn about applications in transition metal complexes, electrochemistry, and lattice optimization using AI. Understand the computation of observables, correlations, and the implementation of algorithms for noisy forces and stresses in quantum materials research.

Auxiliary-field methods for quantum materials
Overview: MC in quantum physics
The many-electron problem
The challenge of accurate calculations
What does AFQMC do?
Connection with other QMC
Understand and control the sign problem
Ligand-dissociation of TM complexes ICTC
Electrochemistry of TM complexes
Computing observables & correlations
Computation of forces and stresses
Lattice optimization - AIN
Optimization with noisy gradients
An algorithm for noisy forces/stresses
Summary