Towards Safe Continuing Task Reinforcement Learning | Read Paper on Bytez