Rare vs. Hard Examples
Dragomir explores the distinction between rare and hard examples in machine learning, emphasizing that not all hard examples are beneficial to label. He discusses the importance of understanding the distribution of examples and introduces concepts like epistemic uncertainty and gradient analysis to better identify long tail scenarios in model training.In this clip
From this podcast

Gradient Dissent - A Machine Learning Podcast
Drago Anguelov — Robustness, Safety, and Scalability at Waymo
Related Questions