Tempognize: Fostering Deep Learning and Knowledge Consolidation with a Large Language Model-Based Learning Assistant
Xin Yu Xie, Rongyu Cui · 2025
The swift progression of artificial intelligence (AI), especially large language models (LLMs), offers both prospects and obstacles for education. Although LLMs provide opportunities for improved knowledge acquisition and individualized learning, they may not intrinsically foster higher-order cognitive abilities. This study presents Tempognize, a novel learning assistant application utilizing LLMs to promote Socratic learning and tackle the widespread problem of "information hoarding." Tempognize seeks to enhance learning and knowledge retention by combining AI-generated questions with innovative note lifecycle management and regular evaluations. An extensive semester-long classroom intervention, evaluated via student course reviews, indicates Tempognize's efficacy in augmenting student engagement, fostering deeper learning, and strengthening critical thinking abilities. This study offers a significant example of the successful incorporation of AI in education.