AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Is It Possible to Train Convolution Networks on a Problem?
Deep syth is a network that just haves log time scales, log time cells. It has the property that if you train it on a problem at somea time scale, and then you training it on a different the same problem, spet up or slowed down, it doesn't care. We can make that without retraining weights. And in any event, you know, if you're building robots, re wot over to go off to the moon, or go off to mars, or going to crawl along the sea floor, or whatever, your have a similar problem with deep net works dealing with gravitational forces.