JustTrainTwice:ImprovingGroupRobustnesswithoutTrainingGroupInformationEvanZheranLiu1BehzadHaghgoo1AnnieS.Chen1AditiRaghunathan1PangWeiKoh1ShioriSagawa1PercyLiang1ChelseaFinn1Abstractcanbeespecially...
ImprovingUltrametricsEmbeddingsThroughCoresetsVincentCohen-Addad1Re´mideJoannisdeVerclos2GuillaumeLagarde3Abstractclusterstructures.Sincethe50s,alargetoolkitofmetricembeddingtechniqueshavebeendeve...
ImprovingMolecularGraphNeuralNetworkExplainabilitywithOrthonormalizationandInducedSparsityRyanHenderson1Djork-Arne´Clevert1FlorianeMontanari1AbstractHowever,theopaquenatureofartificialneuralnetwor...
ImprovingPredictorsviaCombinationAcrossDiverseTaskCategoriesKwangInKim1AbstractPredictorcombinationisrelatedtomulti-tasklearning(MTL)andtransferlearning(TL).However,sincenoaccessPredictorcombinatio...
ImprovingLosslessCompressionRatesviaMonteCarloBits-BackCodingYangjunRuan12KarenUllrich23DanielSevero12JamesTownsend4AshishKhisti1ArnaudDoucet5AlirezaMakhzani12ChrisJ.Maddison12Abstract<latexitsha1_...
ImprovingGradientRegularizationusingComplex-ValuedNeuralNetworksEricYeats1YiranChen1HaiLi1Abstractandthattheseinputsaredifficulttodistinguishfromnor-malinputs.SubsequentinvestigationbyGoodfelloweta...
ImprovingBreadth-WiseBackpropagationinGraphNeuralNetworksHelpsLearningLong-RangeDependenciesDenisLukovnikov1AsjaFischer1Abstract2020),(2)over-smoothing(Lietal.,2018;Chenetal.,2019;Zhao&Akoglu,2020;...
ImprovingGeneralizationinMeta-learningviaTaskAugmentationHuaxiuYao†1Long-KaiHuang2LinjunZhang3YingWei4LiTian2JamesZou1JunzhouHuang2ZhenhuiLi5Abstractquicklylearninganewtaskwithaslittleasafewexampl...
DebiasingModelUpdatesforImprovingPersonalizedFederatedTrainingDurmusAlpEmreAcar1YueZhao2RuizhaoZhu1RamonMatasNavarro2MatthewMattina2PaulN.Whatmough2VenkateshSaligrama1Abstractofdevicesparticipating...
CIFS:ImprovingAdversarialRobustnessofCNNsviaChannel-wiseImportance-basedFeatureSelectionHanshuYan1JingfengZhang2GangNiu2JiashiFeng1VincentY.F.Tan13MasashiSugiyama24Abstracttonomousvehicles(Grigores...
CalibrateBeforeUse:ImprovingFew-ShotPerformanceofLanguageModelsTonyZ.Zhao1EricWallace1ShiFeng2DanKlein1SameerSingh3AbstractInput:Subparacting.Sentiment:NegativeGPT-3canperformnumeroustaskswhenpro-I...
ConViT:ImprovingVisionTransformerswithSoftConvolutionalInductiveBiasesSte´phaned’Ascoli12HugoTouvron2MatthewL.Leavitt2AriS.Morcos2GiulioBiroli12LeventSagun2Abstract1.IntroductionConvolutionalarch...
ImprovingtheSampleandCommunicationComplexityforDecentralizedNon-ConvexOptimization:JointGradientEstimationandTrackingHaoranSun1SongtaoLu2MingyiHong1Abstract1.IntroductionManymodernlarge-scalemachin...
ImprovingTransformerOptimizationThroughBetterInitializationXiaoShiHuang12FelipePe´rez1JimmyBa32MaksimsVolkovs1Abstractetal.,2019;Sunetal.,2019).Despitethebroadapplications,optimizationintheTransfo...
ImprovingtheGatingMechanismofRecurrentNeuralNetworksAlbertGu1CaglarGulcehre2TomPaine2MattHoffman2RazvanPascanu2Abstract(Chungetal.,2014,GRUs).ThesegatedRNNshavebeenverysuccessfulinseveraldifferenta...
ImprovingRobustnessofDeep-Learning-BasedImageReconstructionAnkitRaj1YoramBresler1BoLi2Abstract1.IntroductionDeep-learning-basedmethodsforvariousappli-Adversarialexamplesfordeeplearningbasedmethodsh...
ImprovingMolecularDesignbyStochasticIterativeTargetAugmentationKevinYang1WengongJin2KyleSwanson3ReginaBarzilay2TommiJaakkola2Abstractascomplexgraph-structuredobjects.Asaresult,thesemod-elsrequireco...
ImprovingGeneralizationbyControllingLabel-NoiseInformationinNeuralNetworkWeightsHrayrHarutyunyan1KyleReing1GregVerSteeg1AramGalstyan1Abstractproposednoregularizationdropout0.5weightdecay0.003Inthep...
ASequentialSelfTeachingApproachforImprovingGeneralizationinSoundEventRecognitionAnuragKumar1VamsiKrishnaIthapu1Abstractsoundeventstoproduceexemplars.Theseexemplarsthendrivethetrainingmodulesinsuper...
State-ReificationNetworks:ImprovingGeneralizationbyModelingtheDistributiAonoAfHiddenR10.e0BpreBsentations10.010.0xx10.07.57.55.05.05.05.02.52.50.00.00.00.0AlexLamb1JonathanBinas1AnirudhGoyal2.52.51...