this is really well done!
watched Scott's talk on AI/AGI over a year ago, but this is
-
-
-
thanks :)
Kraj razgovora
Novi razgovor -
-
-
How are RCNs similar/different than RIMs (Recurrent Independent Mechanisms)?
-
I haven’t read that paper yet. Will reply after I read it.
Kraj razgovora
Novi razgovor -
-
-
Dileep some of your work is really commendable
-
Thank you!
Kraj razgovora
Novi razgovor -
-
-
For me it is time to clearly formulate hypothesis about what are the principles of intelligence and move on to prove or disprove them
-
the fact that the DL paradigm aims at understanding an input (e.g. an image) in feedforward way is puzzling. It excludes the iterative process of reasoning, which is the cornerstone of intelligence. Principle 1: Intelligence is an iterative process
Kraj razgovora
Novi razgovor -
-
-
Great talk! Many ideas about HLAI and issues with current approaches nicely articulated. Coincidentally (or really through thinking about this more), I've recently also started associating traditional FFNN as just "amortized inference".
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
I thought this new research was fascinating, kind of puts the whole activation functions we currently have to the testhttps://science.sciencemag.org/content/367/6473/twis …
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.