PinnedInTDS ArchivebyAntoine Labatie·Aug 31, 2019It’s Necessary to Combine Batch Norm and Skip ConnectionsThese techniques must go hand in hand
InTDS ArchivebyAntoine Labatie·Jun 24, 2021Removing Batch Dependence in CNNs by Proxy-Normalizing ActivationsMemory-efficient convolutional neural network training
InTDS ArchivebyAntoine Labatie·Dec 30, 2020The Serendipitous Effectiveness of Weight Decay in Deep LearningComing full circle after unintended change of effects