Header logo is

Dissecting Adam: The Sign, Magnitude and Variance of Stochastic Gradients




The ADAM optimizer is exceedingly popular in the deep learning community. Often it works very well, sometimes it doesn't. Why? We interpret ADAM as a combination of two aspects: for each weight, the update direction is determined by the sign of stochastic gradients, whereas the update magnitude is determined by an estimate of their relative variance. We disentangle these two aspects and analyze them in isolation, gaining insight into the mechanisms underlying ADAM. This analysis also extends recent results on adverse effects of ADAM on generalization, isolating the sign aspect as the problematic one. Transferring the variance adaptation to SGD gives rise to a novel method, completing the practitioner's toolbox for problems where ADAM fails.

Author(s): Lukas Balles and Philipp Hennig
Journal: arXiv preprint arXiv:1705.07774
Year: 2018

Department(s): Probabilistic Numerics
Bibtex Type: Article (article)
Paper Type: Conference

URL: https://arxiv.org/abs/1705.07774


  title = {Dissecting Adam: The Sign, Magnitude and Variance of Stochastic Gradients},
  author = {Balles, Lukas and Hennig, Philipp},
  journal = {arXiv preprint arXiv:1705.07774},
  year = {2018},
  url = {https://arxiv.org/abs/1705.07774}