A Quadratic Actor Network for Model-Free Reinforcement Learning | Read Paper on Bytez