Neural Network Expressibility
Simon discusses the expressibility of neural networks concerning depth and width, highlighting the importance of the function learned during training over network expressibility. Yannic questions the network pathology, revealing insights on the impact of depth on accuracy and the limitations of increasing network size.In this clip
From this podcast
Machine Learning Street Talk (MLST)
#032- Simon Kornblith / GoogleAI - SimCLR and Paper Haul!
Related Questions