Tag
A framework combining deep generative models (flow and score-based) with symbolic regression to automatically infer analytical forms of unnormalized probability densities from samples, demonstrated on toy distributions and lattice physics systems like the XY model and φ⁴ theory renormalization.