AdaVD removes target concepts from diffusion models by soft-projecting value vectors away from the target token direction, with a sigmoid threshold that preserves unrelated prompts.
https : / / huggingface
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Precise, Fast, and Low-cost Concept Erasure in Value Space: Orthogonal Complement Matters
AdaVD removes target concepts from diffusion models by soft-projecting value vectors away from the target token direction, with a sigmoid threshold that preserves unrelated prompts.