SlotMachines:DiscoveringWinningCombinationsofRandomWeightsinNeuralNetworksMaxwellMbabillaAladago1LorenzoTorresani1AbstractLearningtypicallyinvolveseitheroptimizinganetworkfromscratch(Krizhevskyetal...
Double-WinQuant:AggressivelyWinningRobustnessofQuantizedDeepNeuralNetworksviaRandomPrecisionTrainingandInferenceYongganFu1QixuanYu1MengLi2VikasChandra2YingyanLin1AbstractWuetal.,2018).However,DNNs...