A key challenge when deploying reinforcement learning (RL) algorithms in robotics set tings is the large number of interactions between the agent and its environment that are necessary for an optimal behavior to be learned. ...