@inproceedings{zhao-etal-2024-ouroboros,
    title = "Ouroboros: Generating Longer Drafts Phrase by Phrase for Faster Speculative Decoding",
    author = "Zhao, Weilin  and
      Huang, Yuxiang  and
      Han, Xu  and
      Xu, Wang  and
      Xiao, Chaojun  and
      Zhang, Xinrong  and
      Fang, Yewei  and
      Zhang, Kaihuo  and
      Liu, Zhiyuan  and
      Sun, Maosong",
    editor = "Al-Onaizan, Yaser  and
      Bansal, Mohit  and
      Chen, Yun-Nung",
    booktitle = "Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing",
    month = nov,
    year = "2024",
    address = "Miami, Florida, USA",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/fix-sig-urls/2024.emnlp-main.742/",
    doi = "10.18653/v1/2024.emnlp-main.742",
    pages = "13378--13393"
}