Infiniteattention:NNGPandNTKfordeepattentionnetworksJiriHron1YasamanBahri2JaschaSohl-Dickstein2RomanNovak2Abstractetal.,2019;Novaketal.,2019;Li&Liang,2018;Allen-Zhuetal.,2019;Duetal.,2019;Aroraetal...
DualityinRKHSswithInfiniteDimensionalOutputs:ApplicationtoRobustLossesPierreLaforgue1AlexLambert1LucBrogat-Motte1Florenced’Alche´-Buc1Abstractandvariantscopewithdiscretestructures(Tsochantaridise...
InfiniteMixturePrototypesforFew-ShotLearningKelseyR.Allen1EvanShelhamer2HanulShin1JoshuaB.Tenenbaum1AbstractTwopopularclassesofnonparametricmethodsarenearestneighbormethodsandprototypicalmethods.Ne...
GenerativeModelingofInfiniteOccludedObjectsforCompositionalSceneRepresentationJinyangYuan1BinLi1XiangyangXue1Abstractcanautomaticallysegregatethevisualsceneintoconceptualentities(e.g.objects)throug...
LearningInfiniteLayerNetworksWithouttheKernelTrickRoiLivni1DanielCarmon2AmirGloberson2Abstract2014;Danielyetal.,2016;Cho&Saul,2009;Heinemannetal.,2016;Williams,1997).Suchmodelscanbeinter-InfiniteLa...
AnInfiniteHiddenMarkovModelWithSimilarity-BiasedTransitionsColinReimerDawson1ChaofanHuang1ClaytonT.Morrison2Abstractinga“stick”offoftheremainingweightaccordingtoaBeta(1,γ)distribution.Theparamet...