Robustness in AI
Human involvement remains crucial in AI development, especially when considering the robustness of models. If a classification changes with simple alterations, like rotating an image, the questions around explainability become misguided. Focusing first on achieving invariance to basic variables sets a stronger foundation for meaningful explanations in machine learning.In this clip
From this podcast

The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
Invariance, Geometry and Deep Neural Networks with Pavan Turaga - #386 (Video)
Related Questions