1.5k citations · 3.2k across the 15 of their papers we have counts for
Showing 2015Show all
2 papers · 1 filter
cs.LG2015★ 221 cited
Unitary Evolution Recurrent Neural Networks
Martin Arjovsky, Amar Shah, Yoshua Bengio
Recurrent neural networks (RNNs) are notoriously difficult to train. When the eigenvalues of the hidden to hidden weight matrix deviate from absolute value 1, optimization becomes…
math.NA2015★ 1 cited
Saddle-free Hessian-free Optimization
Martin Arjovsky
Nonconvex optimization problems such as the ones in training deep neural networks suffer from a phenomenon called saddle point proliferation. This means that there are a vast numbe…