New benchmark dataset aims to standardize evaluation of AI lesson-generation agents

Researchers introduced LessonBench-V1, a benchmark dataset comprising 647 human-written lessons paired with LLM-based reverse-engineered lesson plans across 240 STEM topics spanning mathematics, physics, chemistry, and computer science. The lessons are drawn from 97 trusted open sources, including LibreTexts and GeeksForGeeks. Each lesson plan is human-reviewed and produced through a pedagogically grounded methodology synthesizing Bloom's Taxonomy, Gagné's Events, Merrill's First Principles, and the 5E Instructional Model. The dataset captures 3,620 learning objectives with pedagogical metadata, enabling systematic and reproducible evaluation of lesson-generation AI agents. The study further proposes a three-dimensional evaluation pipeline for use with the dataset.