SemBench: A Universal Semantic Framework for LLM Evaluation
It addresses the resource-intensive and language-limited problem of LLM evaluation for researchers and practitioners, offering a scalable and cross-lingual solution, though it is incremental as it builds on existing benchmark concepts.
The paper tackles the challenge of evaluating semantic understanding in LLMs by introducing SemBench, a framework that automatically generates synthetic benchmarks using dictionary definitions and a sentence encoder, eliminating the need for curated sentences. Results show strong correlation with standard datasets and stable rankings with few examples.
Recent progress in Natural Language Processing (NLP) has been driven by the emergence of Large Language Models (LLMs), which exhibit remarkable generative and reasoning capabilities. However, despite their success, evaluating the true semantic understanding of these models remains a persistent challenge. Traditional benchmarks such as Word-in-Context (WiC) effectively probe this capability, but their creation is resource-intensive and often limited to high-resource languages. In this paper, we introduce SemBench, a framework for automatically generating synthetic benchmarks that assess the semantic competence of LLMs using only dictionary sense definitions and a sentence encoder. This approach eliminates the need for curated example sentences, making it both scalable and language-independent. We evaluate SemBench in three languages (English, Spanish, and Basque) spanning different levels of linguistic resources, and across a wide range of LLMs. Our results show that rankings derived from SemBench strongly correlate with those obtained from standard WiC datasets. Furthermore, our analysis demonstrates that only a small number of examples is required to achieve stable and meaningful rankings. Overall, SemBench provides a lightweight, adaptable, and data-efficient framework for cross-lingual evaluation of semantic understanding in LLMs.