Entropy-Regularized Reinforcement Learning Explained