W&B Deep Learning Salon - Chirag Agarwal & Charles Frye

This salon will feature: - Chirag Agarwal, post-doctoral fellow at Harvard University, presenting on his recent paper with Sara Hooker, of Google Brain, on using the variance of the gradient to estimate example difficulty (arXiV: ), and - Charles Frye, deep learning educator at Weights & Biases, presenting on a useful test case for developing intuition for machine learning: learning to multiply numbers by 1. This presentation was part of our Virtual Deep Learning Salon!