Towards a theory of learning dynamics in deep state space models

Open in new window