Jurgen Schmidhuber

Jürgen Schmidhuber

A German computer scientist whose lab developed long short-term memory (LSTM) networks in 1997 with Sepp Hochreiter, an architecture that dominated sequence modelling, such as speech and text processing, before the rise of the transformer architecture. Schmidhuber has been an outspoken critic of how credit for early deep learning breakthroughs has been distributed within the field. (See also: Transformer)