SAeUron: Interpretable Concept Unlearning in Diffusion Models with Sparse Autoencoders.

Bartosz Cywinski, Kamil Deja

Browse the full ICML paper archive.