Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Global PIQA

Benchmarks

Task NameDataset NameSOTA ResultTrend
Commonsense MCQGlobal-PIQA
Accuracy80.69
9
Open-ended culturalGlobal-PIQA OE
chrF18.73
9
Multilingual Multiple-Choice ReasoningGlobal PIQA 116 languages 1.0 (test)
Accuracy79.31
6
Showing 3 of 3 rows