How do vision foundation models compare in their ability to understand abstract concepts in images?Answer not yet generated.