AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
The Connection Between Transformers and Fast Weight Algorithms
You're in Schmidhuber I think just today or so had a blog post out connecting transformers to fast weight algorithms. Why has this only taken off so recently and what's the difference? We nearly were able to ask him we were so close to getting him on the podcast he pulled out of the last minute but that's why we're gonna be very nice to you again. Timothy Oh here said in order to solve Francois's arc challenge could transformers be alone enough and then actually an extension of that question from Patriots trader from our discord strange name he's asking about the data efficiency of transformers right because we had the guy on from fair who did the data efficient transform