AI RESEARCH
ICED: Concept-level Machine Unlearning via Interpretable Concept Decomposition
arXiv CS.LG
•
ArXi:2605.14309v1 Announce Type: cross Machine unlearning in Vision-Language Models (VLMs) is typically performed at the image or instance level, making it difficult to precisely remove target knowledge without affecting unrelated semantics. This issue is especially pronounced since a single image often contains multiple entangled concepts, including both target concepts to be forgotten and contextual information that should be preserved.