@misc{indiciae17d7fa318044,
  title = {Pretraining Data Mixtures Enable Narrow Model Selection Capabilities in Transformer Models},
  author = {Steve Yadlowsky and Lyric Doshi and Nilesh Tripuraneni},
  year = {2023},
  url = {https://arxiv.org/abs/2311.00871},
  note = {Source identifier: 2311.00871}
}