Sollicitatievraag bij Multiverse Computing

Could you explain encoder–decoder for Reinforcement Learning?