Fig. 1From: Towards more efficient and robust evaluation of sepsis treatment with deep reinforcement learningOverview of the DIRL-MT modelBack to article page