@inproceedings{garcia-de-herreros-etal-2024-explains,
    title = "What explains the success of cross-modal fine-tuning with {ORCA}?",
    author = "Garcia De Herreros, Paloma  and
      Gautam, Vagrant  and
      Slusallek, Philipp  and
      Klakow, Dietrich  and
      Mosbach, Marius",
    editor = "Tafreshi, Shabnam  and
      Akula, Arjun  and
      Sedoc, Jo{\~a}o  and
      Drozd, Aleksandr  and
      Rogers, Anna  and
      Rumshisky, Anna",
    booktitle = "Proceedings of the Fifth Workshop on Insights from Negative Results in NLP",
    month = jun,
    year = "2024",
    address = "Mexico City, Mexico",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2024.insights-1.2",
    pages = "8--16",
    abstract = "ORCA (Shen et al., 2023) is a recent technique for cross-modal fine-tuning, i.e., applying pre-trained transformer models to modalities beyond their training data. The technique consists primarily of training an embedder and fine-tuning the embedder and model. Despite its high performance on a variety of downstream tasks, we do not understand precisely how each of these components contribute to ORCA{'}s success. Therefore, we run a series of ablations and find that embedder training does not help 2D tasks at all, contrary to what the original paper posits. In 1D tasks, some amount of embedder training is necessary but more is not better. In 4 out of 6 datasets we experiment with, it is model fine-tuning that makes the biggest difference. Through our ablations and baselines, we contribute a better understanding of the individual components of ORCA.",
}