@inproceedings{greco-etal-2022-small,
    title = "A Small but Informed and Diverse Model: The Case of the Multimodal {G}uess{W}hat!? Guessing Game",
    author = "Greco, Claudio  and
      Testoni, Alberto  and
      Bernardi, Raffaella  and
      Frank, Stella",
    editor = "Dobnik, Simon  and
      Grove, Julian  and
      Sayeed, Asad",
    booktitle = "Proceedings of the 2022 CLASP Conference on (Dis)embodiment",
    month = sep,
    year = "2022",
    address = "Gothenburg, Sweden",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2022.clasp-1.1",
    pages = "1--10",
    abstract = "Pre-trained Vision and Language Transformers achieve high performance on downstream tasks due to their ability to transfer representational knowledge accumulated during pretraining on substantial amounts of data. In this paper, we ask whether it is possible to compete with such models using features based on transferred (pre-trained, frozen) representations combined with a lightweight architecture. We take a multimodal guessing task as our testbed, GuessWhat?!. An ensemble of our lightweight model matches the performance of the finetuned pre-trained transformer (LXMERT). An uncertainty analysis of our ensemble shows that the lightweight transferred representations close the data uncertainty gap with LXMERT, while retaining model diversity leading to ensemble boost. We further demonstrate that LXMERT{'}s performance gain is due solely to its extra V{\&}L pretraining rather than because of architectural improvements. These results argue for flexible integration of multiple features and lightweight models as a viable alternative to large, cumbersome, pre-trained models.",
}