Re:Weight-Agnostic Neural Networks
by YeGoblynQueenne on Saturday June 15, 2019 @04:26AM
(#20188945)
Yes, sloppy definition of Turing completenes on my part. I apologise unrservedly.
But, I still don't understand your objection. You agree that a regular automaton is a program, and that it's not Turing complete. So a program does not need to be Turing complete.
I didn't say, nor do I think, that the algorithm described in this article is Turing complete. I think it perfoms program induction. That's as far as I went. I still don't see where Turing completeness comes into it.
But, I still don't understand your objection. You agree that a regular automaton is a program, and that it's not Turing complete. So a program does not need to be Turing complete.
I didn't say, nor do I think, that the algorithm described in this article is Turing complete. I think it perfoms program induction. That's as far as I went. I still don't see where Turing completeness comes into it.