Transformer Innovations
Jürgen discusses the evolution of transformers, highlighting his early contributions in 1991 that laid the groundwork for modern architectures. He explains how predictive coding enabled deep learning by compressing data, and he delves into the mechanics of generative adversarial networks, where two networks compete to enhance their predictive capabilities. This foundational work continues to influence today's AI advancements.In this clip
From this podcast
Machine Learning Street Talk (MLST)
Jurgen Schmidhuber on Humans co-existing with AIs
Related Questions