Benchmarking Linguistic Diversity of Large Language Models
This addresses the issue of preserving linguistic richness in AI-generated content for users and developers concerned about the quality and authenticity of online text.
The paper tackles the problem that large language models (LLMs) often lack human-like linguistic diversity in vocabulary, syntax, and meaning, proposing a framework to benchmark LLMs across these dimensions and analyzing how development choices affect diversity.
The development and evaluation of Large Language Models (LLMs) has primarily focused on their task-solving capabilities, with recent models even surpassing human performance in some areas. However, this focus often neglects whether machine-generated language matches the human level of diversity, in terms of vocabulary choice, syntactic construction, and expression of meaning, raising questions about whether the fundamentals of language generation have been fully addressed. This paper emphasizes the importance of examining the preservation of human linguistic richness by language models, given the concerning surge in online content produced or aided by LLMs. We propose a comprehensive framework for evaluating LLMs from various linguistic diversity perspectives including lexical, syntactic, and semantic dimensions. Using this framework, we benchmark several state-of-the-art LLMs across all diversity dimensions, and conduct an in-depth case study for syntactic diversity. Finally, we analyze how different development and deployment choices impact the linguistic diversity of LLM outputs.