Ten­sor Networks

#SAT with Ten­sor Networks

It turns out that DNNs are re­mark­ably in­ter­pretable.

Graph­i­cal ten­sor no­ta­tion for interpretability