Why Do Better Loss Functions Lead to Less Transferable Features?

Simon Kornblith, Ting Chen, Honglak Lee, Mohammad Norouzi 0002. Why Do Better Loss Functions Lead to Less Transferable Features?. In Marc'Aurelio Ranzato, Alina Beygelzimer, Yann N. Dauphin, Percy Liang, Jennifer Wortman Vaughan, editors, Advances in Neural Information Processing Systems 34: Annual Conference on Neural Information Processing Systems 2021, NeurIPS 2021, December 6-14, 2021, virtual. pages 28648-28662, 2021. [doi]

@inproceedings{KornblithCLN21,
  title = {Why Do Better Loss Functions Lead to Less Transferable Features?},
  author = {Simon Kornblith and Ting Chen and Honglak Lee and Mohammad Norouzi 0002},
  year = {2021},
  url = {https://proceedings.neurips.cc/paper/2021/hash/f0bf4a2da952528910047c31b6c2e951-Abstract.html},
  researchr = {https://researchr.org/publication/KornblithCLN21},
  cites = {0},
  citedby = {0},
  pages = {28648-28662},
  booktitle = {Advances in Neural Information Processing Systems 34: Annual Conference on Neural Information Processing Systems 2021, NeurIPS 2021, December 6-14, 2021, virtual},
  editor = {Marc'Aurelio Ranzato and Alina Beygelzimer and Yann N. Dauphin and Percy Liang and Jennifer Wortman Vaughan},
}