Physically Interpretable Representation Learning with Gaussian Mixture Variational AutoEncoder (GM-VAE)

Tiffany Fan, Murray Cutforth, Marta D'Elia +3

Extracting compact, physically interpretable representations from high-dimensional scientific data is a persistent challenge due to the complex, nonlinear structures inherent in ph…