Benchmarking Multimodal Models for Ukrainian Language Understanding Across Academic and Cultural Domains

TL;DR

This paper introduces ZNO-Vision, a comprehensive multimodal benchmark for Ukrainian language understanding across academic and cultural domains, addressing the lack of evaluation tools for low-resource languages.

Contribution

It presents the first Ukrainian multimodal benchmark derived from university exams and evaluates existing models, highlighting performance gaps and cultural knowledge challenges.

Findings

Few models outperform baseline on the benchmark

Performance drops significantly when translating English benchmarks to Ukrainian

Models show limited understanding of Ukrainian cultural content

Abstract

While the evaluation of multimodal English-centric models is an active area of research with numerous benchmarks, there is a profound lack of benchmarks or evaluation suites for low- and mid-resource languages. We introduce ZNO-Vision, a comprehensive multimodal Ukrainian-centric benchmark derived from standardized university entrance examination (ZNO). The benchmark consists of over 4,300 expert-crafted questions spanning 12 academic disciplines, including mathematics, physics, chemistry, and humanities. We evaluated the performance of both open-source models and API providers, finding that only a handful of models performed above baseline. Alongside the new benchmark, we performed the first evaluation study of multimodal text generation for the Ukrainian language: we measured caption generation quality on the Multi30K-UK dataset, translated the VQA benchmark into Ukrainian, and measured performance degradation relative to original English versions. Lastly, we tested a few models from a cultural perspective on knowledge of national cuisine. We believe our work will advance multimodal generation capabilities for the Ukrainian language and our approach could be useful for other low-resource languages.