SAGE erases concepts from diffusion models by optimizing attack prompts against the model's own text encoder and then fine-tuning that encoder with a global-local retention loss.
In: International Conference on Machine Learning (ICML) (2024) 3, 4, 12
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
SAGE: Exploring the Boundaries of Unsafe Concept Domain with Semantic-Augment Erasing
SAGE erases concepts from diffusion models by optimizing attack prompts against the model's own text encoder and then fine-tuning that encoder with a global-local retention loss.