Neural network wave functions and the sign problem
Attila Szabó, Claudio Castelnovo
Abstract
This paper proposes a neural network architecture and learning protocol to solve the problem of convergence to ground states with a nontrivial sign structure. The authors apply their scheme to conventional antiferromagnets and show that, while it works well on those systems, it is unable to find ground states on frustrated magnets, finding instead low-energy states that exhibit the unfrustrated Marshall sign rule
Topics & Concepts
Sign (mathematics)Artificial neural networkConvergence (economics)Protocol (science)Computer scienceScheme (mathematics)MathematicsArtificial intelligenceAlgorithmSign functionFeature (linguistics)Interval (graph theory)Control theory (sociology)Applied mathematicsFunction (biology)Deep learningTopology (electrical circuits)Quantum many-body systemsMachine Learning in Materials SciencePhysics of Superconductivity and Magnetism