And with the SAC learning algorithm it isn't able to learn how to jump over the hurdle... #reinforcementlearning #robotics #ai