{"url":"/dataset/winogavil","name":"WinoGAViL","full_name":null,"description_markdown":"This dataset is collected via the WinoGAViL game to collect challenging vision-and-language associations. Inspired by the popular card game Codenames, a “spymaster” gives a textual cue related to several visual candidates, and another player has to identify them.\r\n\r\nWe use the game to collect 3.5K instances, finding that they are intuitive for humans (>90% Jaccard index) but challenging for state-of-the-art AI models, where the best model (ViLT) achieves a score of 52%, succeeding mostly where the cue is visually salient.\r\n\r\nResearchers are welcome to evaluate models on this dataset. \r\nA simple intended use is zero-shot prediction:\r\nrun vision-and-language model, producing a score for the (cue,image) pair, and taking the K pairs with the highest scores.\r\n\r\nA supervised setting is also possible, code for re-running the experiments is available in the github repository. https://github.com/WinoGAViL/WinoGAViL-experiments","description_withheld":null,"homepage":"https://winogavil.github.io/","introduced_date":"2022-07-25","introduced_date_note":null,"introduced_by":{"paper":"/paper/winogavil-gamified-association-benchmark-to","title":"WinoGAViL: Gamified Association Benchmark to Challenge Vision-and-Language Models","first_author":"Yonatan Bitton","url":null},"license":{"name":"CC-BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/legalcode"},"modalities":[{"name":"Images","url":"/datasets/modality/images"},{"name":"Texts","url":"/datasets/modality/texts"}],"tasks":[{"name":"Common Sense Reasoning","url":"/task/common-sense-reasoning","datasets_with_task":"/datasets/task/common-sense-reasoning"},{"name":"Visual Reasoning","url":"/task/visual-reasoning","datasets_with_task":"/datasets/task/visual-reasoning"}],"languages":[{"name":"English","url":"/datasets/language/english"}],"variants":["WinoGAViL"],"data_loaders":[{"repo":"https://github.com/huggingface/datasets","url":"https://huggingface.co/datasets/severo/winogavil","frameworks":["tf","pytorch","jax"]},{"repo":"https://github.com/huggingface/datasets","url":"https://huggingface.co/datasets/nlphuji/winogavil","frameworks":["tf","pytorch","jax"]},{"repo":"https://github.com/huggingface/datasets","url":"https://huggingface.co/datasets/yonatanbitton/WinoGAViL","frameworks":["tf","pytorch","jax"]}],"num_papers_in_archive":5,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/visual-reasoning-on-winogavil","task":"Visual Reasoning","dataset_variant":"WinoGAViL","rows":8,"metrics":["Jaccard Index"],"first_row_in_archive_order":{"model":"Humans","paper":"/paper/winogavil-gamified-association-benchmark-to","metrics":{"Jaccard Index":"90"},"code_links":[{"title":"winogavil/winogavil-experiments","url":"https://github.com/winogavil/winogavil-experiments"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/common-sense-reasoning-on-winogavil","task":"Common Sense Reasoning","dataset_variant":"WinoGAViL","rows":1,"metrics":["Jaccard Index"],"first_row_in_archive_order":{"model":"ViLT","paper":"/paper/winogavil-gamified-association-benchmark-to","metrics":{"Jaccard Index":"52"},"code_links":[{"title":"winogavil/winogavil-experiments","url":"https://github.com/winogavil/winogavil-experiments"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/winogavil-gamified-association-benchmark-to","title":"WinoGAViL: Gamified Association Benchmark to Challenge Vision-and-Language Models","date":"2022-07-25","rows_on_this_dataset":9,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":2,"samples_ran":1,"samples_unverified":1,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}}],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":1,"samples_harvested":2,"samples_ran":1,"samples_unverified":1,"pointer_only_for_licence":0,"papers_with_no_sample_that_ran":0,"note":"the per-paper counts above, summed; not a rate"},"papers_note":"The archive never published its papers-using-dataset list; these are papers with a leaderboard row on this dataset's benchmarks."}