Mihály Bányai

I am a staff scientist interested in the theory of representation
learning, especially how humans change their representation as they
learn a task, but also machine learning algorithms that do the same
efficiently. Previously I worked on meta-cognitive reinforcement
learning with Peter Dayan in Tübingen and hierarchical Bayesian models
for the visual cortex with Gergő Orbán at the Wigner Institute.
I’m also very interested in developing better ways to handle software
development in an academic environment, and have been working on
various related efforts.