Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor

This is the paper that introduced the first version of the SAC reinforcement learning algorithm with a constant temperature parameter. Full note: SAC.