@misc{indiciae6b2671be33f2,
  title = {Adding Gradient Noise Improves Learning for Very Deep Networks},
  author = {Arvind Neelakantan and Luke Vilnis and Quoc V. Le and Ilya Sutskever and Lukasz Kaiser and Karol Kurach and James Martens},
  year = {2015},
  url = {https://arxiv.org/abs/1511.06807},
  note = {Source identifier: 1511.06807}
}