RuBLiMP: Russian Benchmark of Linguistic Minimal Pairs

TL;DR

RuBLiMP is a comprehensive Russian linguistic minimal pairs benchmark with 45,000 sentence pairs, designed to evaluate language models' grasp of diverse grammatical phenomena through automated and curated data.

Contribution

It introduces a novel, large-scale Russian minimal pairs benchmark created via linguistic perturbations, expanding evaluation resources for language models.

Findings

Models are sensitive to morphology and agreement contrasts.

Models underperform humans on structural, negation, transitivity, and tense phenomena.

Benchmark is publicly available for further research.

Abstract

Minimal pairs are a well-established approach to evaluating the grammatical knowledge of language models. However, existing resources for minimal pairs address a limited number of languages and lack diversity of language-specific grammatical phenomena. This paper introduces the Russian Benchmark of Linguistic Minimal Pairs (RuBLiMP), which includes 45k pairs of sentences that differ in grammaticality and isolate a morphological, syntactic, or semantic phenomenon. In contrast to existing benchmarks of linguistic minimal pairs, RuBLiMP is created by applying linguistic perturbations to automatically annotated sentences from open text corpora and carefully curating test data. We describe the data collection protocol and present the results of evaluating 25 language models in various scenarios. We find that the widely used language models for Russian are sensitive to morphological and agreement-oriented contrasts but fall behind humans on phenomena requiring understanding of structural relations, negation, transitivity, and tense. RuBLiMP, the codebase, and other materials are publicly available.