While deep reinforcement learning (RL) has been demonstrated effective in solving complex
control tasks, sample efficiency remains a key challenge due to the large amounts of data …