Modelpedia
Work in progress
About
Findings
Models
Concepts
Methods
Datasets
Sources
Light
Dark
Circumventing Concept Erasure Methods For Text-To-Image Generative Models
2024-01-16
· ICLR 2024 poster ·
anchor
Findings
IC-1123
All seven published concept erasure methods applied to Stable Diffusion 1.4 can be circumvented by learned word embeddings, demonstrating that targeted concepts are input-filtered rather than truly removed from the model