Beyond gradients: Tsetlin machines and the logic of learning

Learning without backpropagation, explainable by construction: propositional logic automata and the search for interpretable AI