Iddo Drori, MIT, Columbia University, Boston University
This video discusses using Large Language Models (LLMs) to teach and evaluate machine learning courses by solving, explaining, generating, and grading exam questions. The work analyzes a dataset of exam questions from various courses that are available online and explores LLMs' potential in explaining answers, generating new questions, automatic grading, and generating meta-questions that can be used to evaluate students while allowing the use of LLMs. Results show LLMs pass machine learning finals, and using them in education promotes critical thinking and enhances the learning experience.
