@inproceedings{alvarez-melis-jaakkola-2017-causal,
    title = "A causal framework for explaining the predictions of black-box sequence-to-sequence models",
    author = "Alvarez-Melis, David  and
      Jaakkola, Tommi",
    editor = "Palmer, Martha  and
      Hwa, Rebecca  and
      Riedel, Sebastian",
    booktitle = "Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing",
    month = sep,
    year = "2017",
    address = "Copenhagen, Denmark",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/Add-Cong-Liu-Florida-Atlantic-University-author-id/D17-1042/",
    doi = "10.18653/v1/D17-1042",
    pages = "412--421",
    abstract = "We interpret the predictions of any black-box structured input-structured output model around a specific input-output pair. Our method returns an {\textquotedblleft}explanation{\textquotedblright} consisting of groups of input-output tokens that are causally related. These dependencies are inferred by querying the model with perturbed inputs, generating a graph over tokens from the responses, and solving a partitioning problem to select the most relevant components. We focus the general approach on sequence-to-sequence problems, adopting a variational autoencoder to yield meaningful input perturbations. We test our method across several NLP sequence generation tasks."
}