CONTROVERSIAL STIMULI: PITTING NEURAL NETWORKS AGAINST EACH OTHER AS MODELS OF HUMAN RECOGNITION