CLAMS

Cross-linguistic Analysis of Models on Syntax

TextsApache 2.0Introduced 2020-05-01

Targeted syntactic evaluation datasets in 5 languages: English, French, German, Russian, and Hebrew. Data are translated from the targeted syntactic evaluation data of Marvin & Linzen (2018): https://aclanthology.org/D18-1151/ . All stimuli focus on subject-verb agreement.