can.demircan@helmholtz-munich.de
Can studies the representations of deep learning models, how they relate to humans' representations, and how to reduce these representations to interpretable forms.
