So? I'm still seeing the same object. Large models aren't trained on 10k different images of a single cat.