@inproceedings{sun-hewitt-2023-character,
    title = "Character-Level {C}hinese Backpack Language Models",
    author = "Sun, Hao  and
      Hewitt, John",
    editor = "Belinkov, Yonatan  and
      Hao, Sophie  and
      Jumelet, Jaap  and
      Kim, Najoung  and
      McCarthy, Arya  and
      Mohebbi, Hosein",
    booktitle = "Proceedings of the 6th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP",
    month = dec,
    year = "2023",
    address = "Singapore",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/add-emnlp-2024-awards/2023.blackboxnlp-1.8/",
    doi = "10.18653/v1/2023.blackboxnlp-1.8",
    pages = "106--119",
    abstract = "The Backpack is a Transformer alternative shown to improve interpretability in English language modeling by decomposing predictions into a weighted sum of token sense components. However, Backpacks' reliance on token-defined meaning raises questions as to their potential for languages other than English, a language for which subword tokenization provides a reasonable approximation for lexical items. In this work, we train, evaluate, interpret, and control Backpack language models in character-tokenized Chinese, in which words are often composed of many characters. We find that our (134M parameter) Chinese Backpack language model performs comparably to a (104M parameter) Transformer, and learns rich character-level meanings that log-additively compose to form word meanings. In SimLex-style lexical semantic evaluations, simple averages of Backpack character senses outperform input embeddings from a Transformer. We find that complex multi-character meanings are often formed by using the same per-character sense weights consistently across context. Exploring interpretability-through control, we show that we can localize a source of gender bias in our Backpacks to specific character senses and intervene to reduce the bias."
}