Geometric Deep Learning
Paul discusses the different perspectives on geometric deep learning, highlighting the need to generalize away from invertibility and composability in neural networks. The conversation delves into the abstract concepts of neural networks as morphisms in a two-category and the implications of two-categorical universal algebra in making complex reasoning more tractable.In this clip
From this podcast
Machine Learning Street Talk (MLST)
Dr. Paul Lessard - Categorical/Structured Deep Learning
Related Questions