@article{almostoptimalmodelfreereinforcement,
  title = {Almost Optimal Model-Free Reinforcement Learning via Reference-Advantage Decomposition},
  author = {Zihan Zhang and Yuan Zhou and Xiangyang Ji},
  year = {2020},
  eprint = {2004.10019},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2004.10019v2},
}