BLiMP: The Benchmark of Linguistic Minimal Pairs for English
Alex Warstadt, Alicia Parrish, Haokun Liu, Anhad Mohananey, Wei Peng, Sheng-Fu Wang, Samuel R. Bowman
Code Available — Be the first to reproduce this paper.
ReproduceCode
- github.com/alexwarstadt/data_generationOfficialIn papernone★ 30
- github.com/alexwarstadt/blimpOfficialIn papernone★ 0
- github.com/wwt17/zorronone★ 0
- github.com/phueb/zorronone★ 0
Abstract
We introduce The Benchmark of Linguistic Minimal Pairs (shortened to BLiMP), a challenge set for evaluating what language models (LMs) know about major grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each containing 1000 minimal pairs isolating specific contrasts in syntax, morphology, or semantics. The data is automatically generated according to expert-crafted grammars, and aggregate human agreement with the labels is 96.4%. We use it to evaluate n-gram, LSTM, and Transformer (GPT-2 and Transformer-XL) LMs. We find that state-of-the-art models identify morphological contrasts reliably, but they struggle with semantic restrictions on the distribution of quantifiers and negative polarity items and subtle syntactic phenomena such as extraction islands.