Neural Network Robustness
Hadi and Tim discuss the importance of non-robust features in neural networks, highlighting the significance of testing under various corruptions to ensure model effectiveness. They delve into scenarios where human control over objects of interest can enhance model robustness, using real-life examples like helicopter landing in severe weather conditions.In this clip
From this podcast
Machine Learning Street Talk (MLST)
#52 - Unadversarial Examples (Hadi Salman, MIT)
Related Questions