Invited Talk - Large Language Models for Math (Education): From Problem-Solving to Teaching Problem-Solving

Recent developments in Large Language Models (LLMs) have brought new hope for a transformation of mathematics research and education. LLMs have shown (a) impressive reasoning and problem-solving skills, and (b) a potential to tutor students in their problem-solving activities in mathematics, leading to increased use of LLMs in online learning platforms that serve millions of students around the world. In this talk, I will describe some of my group's effort on (a) improving the mathematical problem-solving abilities of current models, and (b) leveraging LLMs to support mathematics education in various ways, such as creating practice problems and scaffolds, providing meaningful feedback to students, having a Socratic dialog with students, etc. - while customizing the learning experiences of students according to their own individual needs. Finally, I will share some challenges as well as opportunities in this emerging field.