x
FirstPrinciples Talks: Shallow Recurrent Decoders for the Automated Discovery of Physical Models — LessWrong