no code implementations • 29 Apr 2019 • Murphy Yuezhen Niu, Lior Horesh, Isaac Chuang
To understand the fundamental trade-offs between training stability, temporal dynamics and architectural complexity of recurrent neural networks~(RNNs), we directly analyze RNN architectures using numerical methods of ordinary differential equations~(ODEs).
no code implementations • 27 Sep 2018 • Murphy Yuezhen Niu, Lior Horesh, Michael O'Keeffe, Isaac Chuang
We show that most of the existing proposals of RNN architectures belong to different orders of $n$-$t$-ORNNs.