Robust Concept Erasure Using Task Vectors

arXiv:2404.03631v1 Announce Type: new
Abstract: With the rapid growth of text-to-image models, a variety of techniques have been suggested to prevent undesirable image generations. Yet, these methods often only protect against specific user prompts and have been shown to allow unsafe generations with other inputs. Here we focus on unconditionally erasing a concept from a text-to-image model rather than conditioning the erasure on the user's prompt. We first show that compared to input-dependent erasure methods, concept erasure that uses Task …

abstract arxiv concept cs.cv focus growth image inputs prompts protect robust text text-to-image type vectors