@misc{indiciae392330267bb9,
  title = {Escaping mediocrity: how two-layer networks learn hard generalized linear models with SGD},
  author = {Luca Arnaboldi and Florent Krzakala and Bruno Loureiro and Ludovic Stephan},
  year = {2024},
  url = {https://arxiv.org/abs/2305.18502},
  note = {Source identifier: 2305.18502}
}