1 paper · 1 filter
Albert Ierusalem
This paper describes some of the possibilities of artificial neural networks that open up after solving the problem of catastrophic forgetting. A simple model and reinforcement lea…