ParallelizingLegendreMemoryUnitTrainingNarsimhaChilkuri1ChrisEliasmith12Abstractmakeitpossibleforustoexploitresourcessuchasthein-ternet,1whichproduces20TBoftextdataeachmonth.ARecently,anewrecurrent...
OnlineLimitedMemoryNeural-LinearBanditswithLikelihoodMatchingOfirNabati1TomZahavy12ShieMannor13Abstractplorationduringtherepresentationlearningphaseisstillanopenproblem.The-greedypolicy(Langford&Zh...
MemoryEfficientOnlineMetaLearningDurmusAlpEmreAcar1RuizhaoZhu1VenkateshSaligrama1AbstractWhilehumanlearningisevidentlytheinspirationfortradi-tionalmeta-learningapproaches,therearesignificantdif-Wep...
GeneralizableEpisodicMemoryforDeepReinforcementLearningHaoHu1JianingYe2GuangxiangZhu1ZhizhouRen3ChongjieZhang1AbstractDiscreteEpisodicMemoryEpisodicMemory-basedmethodscanrapidlyKeyValuelatchontopas...
EL-Attention:MemoryEfficientLosslessAttentionforGenerationYuYan1JiushengChen1WeizhenQi2NikhilBhendawade1YeyunGong3NanDuan3RuofeiZhang4Abstractpruninglayer(Fanetal.,2019)ortrainingasmallerstu-dentmo...
ActNN:ReducingTrainingMemoryFootprintvia2-BitActivationCompressedTrainingJianfeiChen1LianminZheng1ZheweiYao1DequanWang1IonStoica1MichaelW.Mahoney1JosephE.Gonzalez1AbstractTrainingThroughput60L0L1L3...
Sub-linearMemorySketchesforNearNeighborSearchonStreamingDataBenjaminColeman1RichardGBaraniuk12AnshumaliShrivastava12Abstracthavehighsimilaritytoanydynamicallygeneratedqueryq.Wepresentthefirstsublin...
Self-AttentiveAssociativeMemoryHungLe1TruyenTran1SvethaVenkatesh1Abstractetal.,2012).ItremainsunclearhowthehippocampuscanselectthestoreditemsincleverwaystounearththeirhiddenHeretofore,neuralnetwork...
OptimalContinualLearninghasPerfectMemoryandisNP-HARDJeremiasKnoblauch1HishamHusain2TomDiethe3Abstractsibletore-trainamodeleverytimenewdataarrives.Forexample,datamaybetoosensitiveorexpensivetostoreC...
MultigridNeuralMemoryTriHuynh1MichaelMaire1MatthewR.Walter2Abstractsubsystemwithsuchpropertiescouldbeatransformativetechnology—pushingneuralnetworkswithingraspoftasksWeintroduceanovelapproachthate...
HallucinativeTopologicalMemoryforZero-ShotVisualPlanningKaraLiu1ThanardKurutach1ChristineTung1PieterAbbeel1AvivTamar12Abstracttemdynamicsarenotknown,andonlyadatasetofstatetransitionsisavailable.Inp...
AssociativeMemoryinIteratedOverparameterizedSigmoidAutoencodersYiboJiang1CengizPehlevan2AbstractWestudyauto-encodersinalimitofneuralnetworksthatmakestheoreticalanalysispossible.Specifically,astheRe...
UnderstandingandControllingMemoryinRecurrentNeuralNetworksDoronHaviv12AlexnaderRivkind234OmriBarak23Abstract1.IntroductionTobeeffectiveinsequentialdataprocessing,Re-RecurrentNeuralNetworks(RNN)aret...
ContextualMemoryTreesWenSun1AlinaBeygelzimer2HalDaume´III3JohnLangford3PaulMineiro4Abstractriessoastomaximizethedownstreamrewardofqueries.Inordertoscaletoverylargememories,ourapproachorga-Wedesign...
AStatisticalInvestigationofLongMemoryinLanguageandMusicAlexanderGreaves-Tunnell1ZaidHarchaoui1Abstractundoubtedlyhelpful,suchheuristicsarerarelydefinedwithrespecttoanunderlyingmathematicalorstatist...
APersonalizedAffectiveMemoryModelforImprovingEmotionRecognitionPabloBarros1GermanI.Parisi12StefanWermter1Abstracttechniques(Ngetal.,2015;Kayaetal.,2017),neuralactiva-tionanddatadistributionregulari...
GenerativeTemporalModelswithSpatialMemoryforPartiallyObservedEnvironmentsMarcoFraccaro1DaniloJimenezRezende2YoriZwols2AlexanderPritzel2S.M.AliEslami2FabioViola2Abstractcanbeusefulformodel-basedrein...
FasterDerivative-FreeStochasticAlgorithmforSharedMemoryMachinesBinGu1ZhouyuanHuo1ChengDeng2HengHuang12Abstractsianmatrixoritsapproximations)oftheobjectivefunction.Forexample,Hogwild!(Rechtetal.,201...