erikerikson on NoProp: Training neural networks without back-propagation or forward-propagation · Hackerly

Comment on NoProp: Training neural networks without back-propagation or forward-propagation parent

See also https://en.m.wikipedia.org/w/index.php?title=Generalized_Heb...