Neural Quantum States
Overview
Neural quantum states represent the wavefunction of interacting quantum systems with neural networks, then use variational Monte Carlo to estimate energies and observables without storing every state amplitude. Approaches include autoregressive recurrent networks, transformers, restricted Boltzmann machines, and determinant-based models, applied to spin lattices, strongly correlated electrons, and molecular potential-energy surfaces. Current work improves these models through geometry-aware architectures, parameter sharing across system sizes or molecular geometries, and better gradient estimators, sampling, and optimization. A central tension is that expressive models can capture complex correlations and transfer between systems, but reliable training, unbiased sampling, and guarantees of generalization remain difficult—especially near phase transitions and in topologically nontrivial systems.
Latest papers
27 papers0 papers in the last four weeks, down 100% on the four weeks before. 0.0% of all new papers.