Neural Networks

[Tikib2020]

Drdabbles your being really disingenuous when you say that artificial neural networks aren't like human brains.

Artificial neural networks are sort of a very basic, simulated version of a brain, hence the name. And they learn in a similar way.

Claiming that they are nothing like one another is like claiming dogs and cats are nothing like one another.

Source : Am neuroscience graduate.

 

[DrDabbles]

Drdabbles your being really disingenuous when you say that artificial neural networks aren't like human brains.

Artificial neural networks are sort of a very basic, simulated version of a brain, hence the name. And they learn in a similar way.

Claiming that they are nothing like one another is like claiming dogs and cats are nothing like one another.

Source : Am neuroscience graduate.

You might want to go read a research paper on how an artificial neural network work, then. Because they're not alike in any way, and they absolutely do not learn in the same way at all. Artificial neural networks don't create connections between neurons. Each node on the graph is connected to every node in the previous layer and to every node in the next layer. Honestly. This is ultra basic information that's covered countless times.

 

[heltok]

I called it:

I assume the first step for Tesla is to go from 2.5D to 4D labelling of the dataset, but then I think at some point I think they will change the architecture of their neural network from an inception-style CNN to some form of a transformer style network.

https://twitter.com/karpathy/status/1305302243449516032
https://twitter.com/karpathy/status/1305306119372300290

Andrej Karpathy
@karpathy
Transformers
. Specifically, organizing information processing into multiplicative message passing in graphs; generalizing, simplifying, unifying, improving neural nets across domains. For a while there I was growing bit jaded with slowing progress on neural net architectures
2:28 AM · Sep 14, 2020·Twitter Web App

Replying to
@karpathy
feels like a lot is kicked up in dust, and the closest we've come to a full refactor of your typical neural net. stop me if I'm being overly dramatic