79 citations · 174 across the 7 of their papers we have counts for
Showing stat.MLShow all
2 papers · 1 filter
stat.ML2020★ 9 cited
Batch Normalization Provably Avoids Rank Collapse for Randomly Initialised Deep Networks
Hadi Daneshmand, Jonas Kohler, Francis Bach +2
Randomly initialized neural networks are known to become harder to train with increasing depth, unless architectural enhancements like residual connections and batch normalization…
stat.ML2018★ 28 cited
Exponential convergence rates for Batch Normalization: The power of length-direction decoupling in non-convex optimization
Jonas Kohler, Hadi Daneshmand, Aurelien Lucchi +3
Normalization techniques such as Batch Normalization have been applied successfully for training deep neural networks. Yet, despite its apparent empirical benefits, the reasons beh…