Return to Article Details Terminal Prediction as an Auxiliary Task for Deep Reinforcement Learning Download Download PDF