Neural circuit policies enabling auditable autonomy

Mathias Lechner, Ramin Hasani, Alexander Amini, Thomas A. Henzinger, Daniela Rus, Radu Grosu

DOI: 10.1038/s42256-020-00237-3

Journal: Nature Machine Intelligence

It is discovered that a single algorithm with 19 control neurons, connecting 32 encapsulated input features to outputs by 253 synapses, learns to map high-dimensional inputs into steering commands, showing superior generalizability, interpretability and robustness compared with orders-of-magnitude larger black-box learning systems.