AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Graph Neural Networks - What's Happening?
In an inner graph neral network, you have a fixed number of mess passing up dates. But in a cellular automaturet jus, it continuously happens. And the way that these guys were training the the model was to try and make it converge and not degenerate. So what would a continuous ganin layr look like? Or maybe you don't say layr in that case,. The topology of the under underlying graft as strongly determines where that goes. This is absolutely ating. It really makes me think of a cellular autometercause we interviewed ago, calleda daniel graterola and a heap.