Uhura: A Benchmark for Evaluating Scientific Question Answering and Truthfulness in Low-Resource African Languages
This addresses the problem of limited evaluation resources for low-resource African languages, which is crucial for ensuring safe and reliable AI applications in multilingual contexts, though it is incremental as it builds on existing benchmarks via translation.
The authors tackled the lack of datasets for evaluating large language models on knowledge-intensive tasks in low-resource African languages by creating the Uhura benchmark, which includes multiple-choice science questions and a truthfulness test across six languages, revealing a significant performance gap between proprietary and open-source models and poorer performance in African languages compared to English.
Evaluations of Large Language Models (LLMs) on knowledge-intensive tasks and factual accuracy often focus on high-resource languages primarily because datasets for low-resource languages (LRLs) are scarce. In this paper, we present Uhura -- a new benchmark that focuses on two tasks in six typologically-diverse African languages, created via human translation of existing English benchmarks. The first dataset, Uhura-ARC-Easy, is composed of multiple-choice science questions. The second, Uhura-TruthfulQA, is a safety benchmark testing the truthfulness of models on topics including health, law, finance, and politics. We highlight the challenges creating benchmarks with highly technical content for LRLs and outline mitigation strategies. Our evaluation reveals a significant performance gap between proprietary models such as GPT-4o and o1-preview, and Claude models, and open-source models like Meta's LLaMA and Google's Gemma. Additionally, all models perform better in English than in African languages. These results indicate that LMs struggle with answering scientific questions and are more prone to generating false claims in low-resource African languages. Our findings underscore the necessity for continuous improvement of multilingual LM capabilities in LRL settings to ensure safe and reliable use in real-world contexts. We open-source the Uhura Benchmark and Uhura Platform to foster further research and development in NLP for LRLs.