I don't think anyone actually believes the brain backpropates, I think the closest is Hinton saying that the error signal is very useful and the brain uses it somehow
-
-
I guess it breaks into 2 cases: 1) you don’t need the gradient, or 2) you do but you have a way to get it that is structurally very different than backprop. For instance this paper gets gradient w/ either backprop or EM (perhaps an example of case #2): https://arxiv.org/abs/1202.3732
-
Also I think "the brain minimizes an objective function" is a vacuous statement. All dynamical systems can be reframed as solving a variational problem.
- 15 more replies
New conversation -
-
-
I was more commenting on the sociology of the field, where CoSyNe tends to pick up on whatever was big in ML a few years ago.
-
And I think the reason neurotheory tends to go in cycles based on ML trends rather than building on prior work is the lack of experimental validation.
- 2 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
