Task-OrientedActivePerceptionandPlanninginEnvironmentswithPartiallyKnownSemanticsMahsaGhasemi1ErdemArincBulgur2UfukTopcu2AbstractJointperceptionandplanningposetwofundamentalchal-lenges.Thefirstchal...
FastAdaptationtoNewEnvironmentsviaPolicy-DynamicsValueFunctionsRobertaRaileanu1MaxGoldstein1ArthurSzlam2RobFergus1Abstractmighthavetoadjustitsbehaviordependingonweathercon-ditions,oraprostheticcont...
ActiveWorldModelLearninginAgent-richEnvironmentswithProgressCuriosityKunoKim1MegumiSano1JulianDeFreitas2NickHaber3DanielYamins14Abstractmotionlessball,yougrowbored.Youconsiderthemerry-go-roundabitm...
RegularizationinDirectableEnvironmentswithApplicationtoTetrisJanMalteLichtenberg1O¨zgu¨rS¸ims¸ek1Abstractonregularization.Specifically,weproposeamodelthatintroducesabiastowardgivingallfeaturese...
ReinforcementLearninginConfigurableContinuousEnvironmentsAlbertoMariaMetelli1EmanueleGhelfi1MarcelloRestelli1AbstractasaConfigurableMarkovDecisionProcess(Conf-MDP,Metellietal.,2018).Asintraditional...
CausalDiscoveryandForecastinginNonstationaryEnvironmentswithState-SpaceModelsBiweiHuang1KunZhang1MingmingGong12ClarkGlymour1Abstractnonstationarytimeseries,andconcernedwithbothfindingcausalrelation...