Framework measures how naturally LLMs match human language patterns

Original: How Human-Like Are Large Language Models? A Register-Aware Linguistic Evaluation Framework

Source: arxiv.org ↗