FOP:FactorizingOptimalJointPolicyofMaximum-EntropyMulti-AgentReinforcementLearningTianhaoZhang1YuehengLi1ChenWang1GuangmingXie1ZongqingLu1Abstractvalue-basedandactor-criticMARLmethods,whereglobalin...
DisambiguationofWeakSupervisionleadingtoExponentialConvergenceratesVivienCabannes1FrancisBach1AlessandroRudi1Abstractsupervisedlearningareusuallybasedoncleverheuristics,andtheoreticalfoundationsofl...
Differentially-PrivateClusteringofEasyInstancesEdithCohen12HaimKaplan12YishayMansour12UriStemmer13EliadTsfadia12Abstractandmanydifferentalgorithmshavebeensuggested.1How-ever,tothebestofourknowledge...
DetectionofSignalintheSpikedRectangularModelsJiHyungJung1HyeWonChung2JiOonLee13Abstractv∈RNwithu=v=1.Theparameterλcorrespondstothesignal-to-noiseratio(SNR).Weconsidertheproblemofdetectingsignalsi...
DenseforthePriceofSparse:ImprovedPerformanceofSparselyInitializedNetworksviaaSubspaceoffsetIlanPrice12JaredTanner12Abstract1.1.CompetingPrioritiesforSparseNetworksThatneuralnetworksmaybeprunedtohig...
Data-drivenPredictionofGeneralHamiltonianDynamicsviaLearningExactly-SymplecticMapsRenyiChen1MoleiTao1Abstract1.IntroductionWeconsiderthelearningandpredictionofnon-Givenacollectionofsequences,eachbe...
CumulantsofHawkesProcessesareRobusttoObservationNoiseWilliamTrouleau1JalalEtesami2MatthiasGrossglauser1NegarKiyavash2PatrickThiran1Abstracttheyareusedtomodelthestochastictimeevolutionoflimitorderbo...
CountSketches,FeatureHashingandtheMedianofThreeKasperGreenLarsen1RasmusPagh2JakubTeˇtek2Abstractmultitasklearning(Weinbergeretal.,2009)—seesec-tion1.5formoredetails.Inthispaper,werevisittheclassi...
Double-WinQuant:AggressivelyWinningRobustnessofQuantizedDeepNeuralNetworksviaRandomPrecisionTrainingandInferenceYongganFu1QixuanYu1MengLi2VikasChandra2YingyanLin1AbstractWuetal.,2018).However,DNNs...
ConcentricMixturesofMallowsModelsforTop-kRankings:SamplingandIdentifiabilityFabienCollas1EkhineIrurozki12Abstractmillionsofresults.Inthispaper,westudymixturesoftwoMallowsTheMallowsmodel(MM)(Fligner...
CLOCS:ContrastiveLearningofCardiacSignalsAcrossSpace,Time,andPatientsDaniKiyasseh1TingtingZhu1DavidA.Clifton12Abstractsharesomecontext.Bycapturingthishigh-levelsharedcontext(e.g.,medicaldiagnosis),...
CIFS:ImprovingAdversarialRobustnessofCNNsviaChannel-wiseImportance-basedFeatureSelectionHanshuYan1JingfengZhang2GangNiu2JiashiFeng1VincentY.F.Tan13MasashiSugiyama24Abstracttonomousvehicles(Grigores...
CharacterizingStructuralRegularitiesofLabeledDatainOverparameterizedModelsZihengJiang123ChiyuanZhang4KunalTalwar45MichaelC.Mozer46Abstract(KISS!KISSED,KICK!KICKED,BREW!BREWED,etc.).Generalizationto...
CharacterizingFairnessOvertheSetofGoodModelsUnderSelectiveLabelsAmandaCoston1AsheshRambachan2AlexandraChouldechova3Abstractousmethodsforlearninganewthebestperformingmodelamongthosethatsatisfyachose...
CalibrateBeforeUse:ImprovingFew-ShotPerformanceofLanguageModelsTonyZ.Zhao1EricWallace1ShiFeng2DanKlein1SameerSingh3AbstractInput:Subparacting.Sentiment:NegativeGPT-3canperformnumeroustaskswhenpro-I...
BreakingtheLimitsofMessagePassingGraphNeuralNetworksMuhammetBalcilar12PierreHe´roux1BenoitGau¨ze`re3PascalVasseur14Se´bastienAdam1PaulHoneine1Abstractableweights.Theseweightscanbesharedwithrespe...
BayesianAlgorithmExecution:EstimatingComputablePropertiesofBlack-boxFunctionsUsingMutualInformationWillieNeiswanger1KeAlexanderWang1StefanoErmon1AbstracttackledbyBayesianoptimizationmethods(Shahria...
BASELayers:SimplifyingTrainingofLarge,SparseModelsMikeLewis1ShrutiBhosale1TimDettmers12NamanGoyal1LukeZettlemoyer12AbstractWorker1Worker2Weintroduceanewbalancedassignmentofex-Re-routetooriginalwork...
BackpropagatedNeighborhoodAggregationforAccurateTrainingofSpikingNeuralNetworksYukunYang1WenruiZhang1PengLi1Abstractetal.,2014;Daviesetal.,2018).WhilerecentSNNresearchadvanceshaveledtoimprovedperfo...
BoostingtheThroughputandAcceleratorUtilizationofSpecializedCNNInferenceBeyondIncreasingBatchSizeJackKosaian1AmarPhanishayee2MatthaiPhilipose2DebadeeptaDey2K.V.Rashmi1AbstractHsiehetal.,2018;Kangeta...