Hatsune Ichidate

Evaluating Readability of LLM-Simplified Texts Using Human Judgments and Automated Metrics

It is suggested that evaluating LLM-based readability adaptation should combine objective measures with learner-centered feedback, rather than relying solely on conventional text difficulty.