@inproceedings{zhao-bethard-2020-berts,
    title = "How does {BERT}`s attention change when you fine-tune? An analysis methodology and a case study in negation scope",
    author = "Zhao, Yiyun  and
      Bethard, Steven",
    editor = "Jurafsky, Dan  and
      Chai, Joyce  and
      Schluter, Natalie  and
      Tetreault, Joel",
    booktitle = "Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics",
    month = jul,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/jlcl-multiple-ingestion/2020.acl-main.429/",
    doi = "10.18653/v1/2020.acl-main.429",
    pages = "4729--4747",
    abstract = "Large pretrained language models like BERT, after fine-tuning to a downstream task, have achieved high performance on a variety of NLP problems. Yet explaining their decisions is difficult despite recent work probing their internal representations. We propose a procedure and analysis methods that take a hypothesis of how a transformer-based model might encode a linguistic phenomenon, and test the validity of that hypothesis based on a comparison between knowledge-related downstream tasks with downstream control tasks, and measurement of cross-dataset consistency. We apply this methodology to test BERT and RoBERTa on a hypothesis that some attention heads will consistently attend from a word in negation scope to the negation cue. We find that after fine-tuning BERT and RoBERTa on a negation scope task, the average attention head improves its sensitivity to negation and its attention consistency across negation datasets compared to the pre-trained models. However, only the base models (not the large models) improve compared to a control task, indicating there is evidence for a shallow encoding of negation only in the base models."
}