@misc{huben2023attention,
  title={Attention-Only Transformers and Implementing MLPs with Attention Heads},
  author={Robert Huben and Valerie Morris},
  year={2023},
  eprint={2309.08593},
  archivePrefix={arXiv},
  primaryClass={cs.LG},
  url={https://arxiv.org/abs/2309.08593}
}