TowardsAdaptiveResidualNetworkTraining:ANeural-ODEPerspectiveChengyuDong1LiyuanLiu2ZichaoLi1JingboShang1Abstractisnotnew(Changetal.,2017;Wenetal.,2019),butthedynamicsofgrowingispoorlyunderstood.Inp...
TowardsAccuratePost-trainingNetworkQuantizationviaBit-SplitandStitchingPeisongWang1QiangChen1XiangyuHe1JianCheng1AbstracthardwarelikeTPUorgeneralhardwarelikeCPUandGPU.Byturningthefloating-pointvalu...
ProperNetworkInterpretabilityHelpsAdversarialRobustnessinClassificationAkhilanBoopathy1SijiaLiu2GaoyuanZhang2CynthiaLiu1Pin-YuChen2ShiyuChang2LucaDaniel1Abstractdefensivedistillation(Papernotetal.,...
OntheTheoreticalPropertiesoftheNetworkJackknifeQiaohuiLin1RobertLunde1PurnamritaSarkar1Abstractetal.,2017;Gonenetal.,2010;Kallaugheretal.,2019).However,comparativelylittleattentionhasbeenpaidtoWest...
NeuralNetworkControlPolicyVerificationwithPersistentAdversarialPerturbationsYuh-ShyangWang1Tsui-WeiWeng2LucaDaniel2AbstractneuralNetworksaresurprisinglyvulnerabletoadversarialexamplesandattacks(Hua...
GoodSubNetworksProvablyExist:PruningviaGreedyForwardSelectionMaoYe1ChengyueGong1LizhenNie2DennyZhou3AdamKlivans1QiangLiu1Abstractachievinghuman-levelaccuracyoncomplexcognitivetasks,includingimagecl...
MulticlassNeuralNetworkMinimizationviaTropicalNewtonPolytopeApproximationGeorgiosSmyrnis1PetrosMaragos1AbstracttherecentadventofneuralNetworkshasprovideditwithan-otherapplicationofgreatinterest,whi...
Multi-AgentRoutingValueIterationNetworkQuinlanSykoraMengyeRenRaquelUrtasunAbstractFigure1.Avisualizationoftherouteproducedbyafleetoftwentyvehiclesusingourproposedalgorithm.ColorsdenotedifferentInth...
LorentzGroupEquivariantNeuralNetworkforParticlePhysicsAlexanderBogatskiy1BrandonAnderson23JanT.Offermann1MarwahRoussi1DavidW.Miller14RisiKondor256Abstractproblemsmayrequireorbenefitfromlatentspacer...
GraphConvolutionalNetworkforRecommendationwithLow-passCollaborativeFiltersWenhuiYu1ZhengQin12Abstractthesignalofeachnodewithitsneighborhoods(Zhouetal.,2018).Byconstrainingthekernel,GCNslosetheabili...
GenerativeAdversarialImitationLearningwithNeuralNetworkParameterization:GlobalOptimalityandConvergenceRateYufengZhang1QiCai1ZhuoranYang2ZhaoranWang1Abstractoptimalpolicy.IRLformulatesILasabilevelop...
EvolutionaryTopologySearchforTensorNetworkDecompositionChaoLi1ZhunSun1AbstractMorerecently,somestudiesbroughttheideafromphysicstoextendTDintoamoresophisticatedform,i.e.,tensornet-TensorNetwork(TN)d...
EstimatingModelUncertaintyofNeuralNetworksinSparseInformationFormJongseokLee1MatthiasHumt1JianxiangFeng12RudolphTriebel12Abstractferencewhatlinearmapsareforalgebra.Forexample,oncetheposteriordistri...
DropNet:ReducingNeuralNetworkComplexityviaIterativePruningJohnTanChongMin1MehulMotani1Abstractodsofreducingcomplexityincludequantizationto16-bit(Guptaetal.,2015),groupL1orL2regularization(AlemuMode...
DistributedOnlineOptimizationoveraHeterogeneousNetworkNimaEshraghi1BenLiang1Abstract2013),Networking(Shietal.,2018),andmachinelearning(Shalev-Shwartz,2012).Indistributedonlineoptimizationoveracom-p...
DeBayes:aBayesianMethodforDebiasingNetworkEmbeddingsMaartenBuyl1TijlDeBie1Abstractavoidingwhatisseenasdisparatetreatment.Yetmoralstandardsandanti-discriminationlawsaretypicallyalsoAsmachinelearning...
BoostingDeepNeuralNetworkEfficiencywithDual-ModuleInferenceLiuLiu12LeiDeng2ZhaodongChen2YukeWang1ShuangchenLi3JingweiZhang3YihuaYang3ZhenyuGu3YufeiDing1YuanXie2Abstract2013;Heetal.,2019;Wangetal.,2...
TheImpactofNeuralNetworkOverparameterizationonGradientConfusionandStochasticGradientDescentKarthikA.Sankararaman12SohamDe3ZhengXu2W.RonnyHuang2TomGoldstein2AbstractClassicalstochasticoptimizationth...
AnExplicitlyRelationalNeuralNetworkArchitectureMurrayShanahan12KyriacosNikiforou1AntoniaCreswell1ChristosKaplanis1DavidBarrett1MartaGarnelo1AbstractRepresentationsthataregeneral-purposeandreusablei...
AFinite-TimeAnalysisofQ-LearningwithNeuralNetworkFunctionApproximationPanXu1QuanquanGu1AbstractwhichtriggersalineofresearchondeepreinforcementlearningsuchasDoubleDeepQ-Learning(VanHasseltQ-learning...