Episode 32: Jamie Simon, UC Berkeley: On theoretical principles for how neural networks learn and generalize
Generally Intelligent
00:00
The Importance of Qualitative Agreement in Neural Networks
The study of infinite width is a very valuable foothold for people who are trying to understand how to think about neural networks. When you say there's qualitative agreement between a thousand wide network versus an infinite with network What kind of qualitative agreement? Yeah, so to answer that I think we need to step back and Try to think about what it is we're trying to do when you understand neural networks This is a deep question that often gets swept under the rock and I think merits more discussion.
Transcript
Play full episode
