@inproceedings{wang-etal-2023-treepiece,
    title = "Treepiece: Faster Semantic Parsing via Tree Tokenization",
    author = "Wang, Sid  and
      Shrivastava, Akshat  and
      Livshits, Aleksandr",
    editor = "Bouamor, Houda  and
      Pino, Juan  and
      Bali, Kalika",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2023",
    month = dec,
    year = "2023",
    address = "Singapore",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-emnlp/2023.findings-emnlp.740/",
    doi = "10.18653/v1/2023.findings-emnlp.740",
    pages = "11082--11092",
    abstract = "\textit{Autoregressive} (AR) encoder-decoder neural networks have proved successful in many NLP problems, including \textit{Semantic Parsing} {--} a task that translates natural language to machine-readable \textit{parse trees}. However, the sequential prediction process of AR models can be slow. To accelerate AR for semantic parsing, we introduce a new technique called \textit{TreePiece} that tokenizes a parse tree into subtrees and generates one subtree per decoding step. On TOPv2 benchmark, TreePiece shows 4.6 times faster decoding speed than standard AR, and comparable speed but significantly higher accuracy compared to \textit{Non-Autoregressive} (NAR)."
}