TrajectoryDiversityforZero-ShotCoordinationAndreiLupu1BrandonCui2HengyuanHu2JakobFoerster2AbstractFigure1:Thecorridorcoordinationtaskadmitsthreeopti-malSPpolicies,onlyoneofwhichissuitableforZSC.Wes...
ModellingBehaviouralDiversityforLearninginOpen-EndedGamesNicolasPerezNieves12YaodongYang13OliverSlumbers3DavidHenryMguni1YingWen3JunWang13AbstractdevelopingAIsthatachievesuper-humanperformanceonsop...
DiversityActor-Critic:Sample-AwareEntropyRegularizationforSample-EfficientExplorationSeungyulHan1YoungchulSung1Abstractforchallengingcontinuouscontroltasks.Inthispaper,sample-awarepolicyentropyregu...
ReliableFidelityandDiversityMetricsforGenerativeModelsMuhammadFerjadNaeem12SeongJoonOh1YoungjungUh1YunjeyChoi1JaejunYoo13AbstractNonetheless,theremustbesomemeasurementofthequalityofgenerativemodels...
BeyondSignalPropagation:IsFeatureDiversityNecessaryinDeepNeuralNetworkInitialization?YanivBlumenfeld1DarGilboa2DanielSoudry1Abstractadetailedanalysisoftheirdynamicsandgeneralizationproperties.Unlik...
ImprovingAdversarialRobustnessviaPromotingEnsembleDiversityTianyuPang1KunXu1ChaoDu1NingChen1JunZhu1Abstract2017b;Athalyeetal.,2018),whereimageswithhumanimperceptibleperturbations(i.e.,adversarialex...