← Back to UX News Feed
Towards demystifying the creativity of diffusion models
We show that a diffusion model’s creativity (its ability to generate novel data, rather than just memorize its training set) is a mathematical consequence of neural networks learning a “smoothed” version of the score function, driving the model to interpolate between training data points along the hidden data manifold.
Full Product UX article at Google Research »
Fair use excerpts with source attribution for comment, news reporting and instructive commentary only. Original summary description and analysis by UXdesign.com’s authors. Original content © Google Research.