نسخة أولية وصول مفتوح
Metonymic Circuits for Abstract Concept Grounding in Vision Transformers
We study how Vision Transformers ground abstract concepts (e.g., angry) when training data provide limited direct referential evidence. We hypothesize a metonymic grounding mechanism in which abstract predictions are driven by concrete, interpretable anchor concepts (e.g., fire) that bridge visual signals to abstract s …