Learning the generating functional for variance reduction in lattice QCD
Researchers have developed a new methodology using machine-learned normalizing flows to reduce variance in lattice gauge field theory calculations. This approach encodes the generating functional, enabling the systematic creation of noiseless estimators for correlation functions. The technique was demonstrated on Quantum Chromodynamics and Yang-Mills theory, achieving up to a three-orders-of-magnitude reduction in variance for glueball correlation functions and Wilson loops. AI