Poster
in
Workshop: 2nd Workshop on Generative AI and Law (GenLaw ’24)
Memorization is Localized within a Small Subspace in Diffusion Models
Ruchika Chavhan · Yongshuo Zong · Ondrej Bohdal · Da Li · Timothy Hospedales
Abstract:
We present an intriguing discovery - Memorization resides within a potentially unique compact subspace in pre-trained models and demonstrate that simply pruning these subspaces leads to effective mitigation of memorization. This marks the first instance of mitigating memorization in diffusion models without additional training, opening new doors for more trustworthy text-to-image generation.
Chat is not available.