Is this the math generalization of multilayer perceptrons? With hidden variables ~= hidden layers?
-
-
-
Interesting idea. But I don't think so in any useful sense, I guess there's no backpropagation and no gradient descent on parameters (unless there is, I didn't read the thing yet)
- Još 2 druga odgovora
Novi razgovor -
-
-
I'm obsessed with this paper. Composing functions by literally drawing pictures of their graphs and doing matrix multiplication.. it sounds like something from a fever dream.
- Još 1 odgovor
Novi razgovor -
-
-
This paper is nuts. I've spent awhile obsessing over it. Does it help your research at all?
-
No idea, I need to read it first. Probably not directly
Kraj razgovora
Novi razgovor -
-
-
Oh hey, this is cool!
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
This is a *very* good find. Thanks so much for sharing!!!!
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.
in