PuzzleMix:ExploitingSaliencyandLocalStatisticsforOptimalMixupJang-HyunKim12000Choo12HyunOhSong12Wo0n0h1o002003Abstract004Input1InputMixupPuzzleMix(zonly)Whiledeepneuralnetworksachievegreatper-005fo...
Mix&Match–AgentCurriculaforReinforcementLearningWojciechMarianCzarnecki1SiddhantM.Jayakumar1MaxJaderberg1LeonardHasenclever1YeeWhyeTeh1SimonOsindero1NicolasHeess1RazvanPascanu1AbstractFigure1.Sche...