DiversityActor-Critic:Sample-AwareEntropyRegularizationforSample-EfficientExplorationSeungyulHan1YoungchulSung1Abstractforchallengingcontinuouscontroltasks.Inthispaper,sample-awarepolicyentropyregu...
ActiveTesting:Sample–EfficientModelEvaluationJannikKossen1SebastianFarquhar1YarinGal1TomRainforth2AbstractDifferencetoFullTestLoss×10−2I.I.D.Acquisition5ActiveTestingWeintroduceanewframeworkfors...
Dimension-WiseImportanceSamplingWeightClippingforSample-EfficientReinforcementLearningSeungyulHan1YoungchulSung1Abstractsamplesgeneratedbythebehaviorpolicywhichcanbedif-ferentfromthetargetpolicy.Of...