@misc{indiciae19fcdcf8b100,
  title = {Never Train from Scratch: Fair Comparison of Long-Sequence Models Requires Data-Driven Priors},
  author = {Ido Amos and Jonathan Berant and Ankit Gupta},
  year = {2024},
  url = {https://arxiv.org/abs/2310.02980},
  note = {Source identifier: 2310.02980}
}