67 Sub-Datasets × 1,000 Minimal Pairs

BLiMP

154
306
CC-BY-4.0
10K – 100K
Text Classification
English

Opened about a year ago

by @yuro-1981

0
@yuro-1981a year ago

BLiMP offers 67 sub-datasets, each with 1,000 minimal pairs, making it incredibly thorough for linguistic evaluation.


0/10000

Sign up or Log in to comment