“one learning algorithm” hypothesis
ExpandDescendants
Properties