1. Wide neural networks of any depth evolve as linear models under gradient descent*This article is an updated version of a paper presented at 33rd Conference on Neural Information Processing Systems (NeurIPS 2019), Vancouver, Canada. (21st December 2020) Authors: Lee, Jaehoon; Xiao, Lechao; Schoenholz, Samuel S; Bahri, Yasaman; Novak, Roman; Sohl-Dickstein, Jascha; Pennington, Jeffrey Journal: Journal of statistical mechanics Issue: (2020:Dec.) Page Start: Record Type: Journal Article View Content: Available online (eLD content is only available in our Reading Rooms) ↗