@misc{indiciae5517ea3e7054,
  title = {Taylorized Training: Towards Better Approximation of Neural Network Training at Finite Width},
  author = {Yu Bai and Ben Krause and Huan Wang and Caiming Xiong and Richard Socher},
  year = {2020},
  url = {https://arxiv.org/abs/2002.04010},
  note = {Source identifier: 2002.04010}
}