Wide Neural Networks of Any Depth Evolve as Linear Models Under Gradient Descent
Jaehoon Lee, Lechao Xiao, Samuel Schoenholz, Yasaman Bahri, Roman Novak, Jascha Sohl-Dickstein, Jeffrey Pennington
–Neural Information Processing Systems
However, the often complex loss landscapes of neural networks have made a theory of learning dynamics elusive.
Neural Information Processing Systems
Oct-2-2025, 01:31:12 GMT