@davidholzmller

Training two-layer ReLU networks with gradient descent is inconsistent

, and . arXiv:2002.04861, (2020)Published at https://jmlr.org/papers/v23/20-830.html.

Links and resources

Tags

community

  • @unibiblio
  • @simtech
  • @davidholzmller
  • @katharinafuchs
  • @mathematik
@davidholzmller's tags highlighted