AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
The Constraints of Neural Networks
The success of neural networks over the last 20 years has been around let's make really big data sets. And we can scale the networks accordingly, even though we're not sort of changing what we put in and out that works the other way too in terms of edge stuff. So are there different kinds of models? Are they the same models that you would use in a large system? Or do you have to constrain them in some way? Yeah, so constrained is the is the underlying word there.