@inproceedings{kerinec-etal-2018-deep,
    title = "When does deep multi-task learning work for loosely related document classification tasks?",
    author = "Kerinec, Emma  and
      Braud, Chlo{\'e}  and
      S{\o}gaard, Anders",
    editor = "Linzen, Tal  and
      Chrupa{\l}a, Grzegorz  and
      Alishahi, Afra",
    booktitle = "Proceedings of the 2018 {EMNLP} Workshop {B}lackbox{NLP}: Analyzing and Interpreting Neural Networks for {NLP}",
    month = nov,
    year = "2018",
    address = "Brussels, Belgium",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-emnlp/W18-5401/",
    doi = "10.18653/v1/W18-5401",
    pages = "1--8",
    abstract = "This work aims to contribute to our understanding of \textit{when} multi-task learning through parameter sharing in deep neural networks leads to improvements over single-task learning. We focus on the setting of learning from \textit{loosely related} tasks, for which no theoretical guarantees exist. We therefore approach the question empirically, studying which properties of datasets and single-task learning characteristics correlate with improvements from multi-task learning. We are the first to study this in a text classification setting and across more than 500 different task pairs."
}