S3E6: Geoffrey Hinton on the Exploration of Thought
Theory and Practice
00:00
The So-Called Neural Tangent
Kolonel methods are about mapping data into a space that's infinite dimensional. Neral networks do the exact opposite, starting with same dimensionality and then mapping it now to a much smaller dimension. These two very kind of different schools of thought actually ended up kind of converging on the same mathematical structure. And so i think this is still the dust hasn't settled, and there's a lot of back and forth in machine learning right now. I'm detion tolearn more about that. But it does t the rubber will meet the road when it comes down to the representations that we learn.
Transcript
Play full episode
