KnowSim: Evaluating Information Calibration in LLM Assistants with User Simulators that Learn

Yoonjoo LeeHyoungwook JinTae Soo KimShaoyang ZhangPhilippe LabanQ. Vera Liao

RelianceScope: An Analytical Framework for Examining Students' Reliance on Generative AI Chatbots in Problem Solving

Hyoungwook JinMinju YooJieun HanZixin ChenSo Yeon AhnXu Wang

When Scaffolding Breaks: Investigating Student Interaction with LLM-Based Writing Support in Real-Time K-12 EFL Classrooms

Junho MyungHyunseung LimHana OhHyoungwook JinNayeon KangSo Yeon AhnHwajung HongAlice OhJuho Kim

TeachTune: Reviewing Pedagogical Agents Against Diverse Student Profiles with Simulated Students

Hyoungwook JinMinju YooJeongeon ParkYokyung LeeXu WangJuho Kim

When to Give Feedback: Exploring Tradeoffs in the Timing of Design Feedback

Jane L EGrace Yu-Chun YenIsabelle Yan PanGrace LinMingyi LiHyoungwook JinMengyi ChenHaijun XiaSteven P. Dow

Teach AI How to Code: Using Large Language Models as Teachable Agents for Programming Education

Hyoungwook JinSeonghee LeeHyungyu ShinJuho Kim

ProcessGallery: An Online Gallery that Highlights Improvements by Principles through Contrasting Pairs of Examples

Grace Yu-Chun YenJane L EHyoungwook JinMingyi LiGrace LinIsabelle Yan PanSteven P. Dow

CodeTree: Learnersourcing Subgoal Hierarchies in Code Examples

Hyoungwook JinJuho Kim

Students increasingly turn to AI for help with problem-solving, yet they struggle to formulate targeted questions. We developed HelpCoach, an add-on for chat interfaces that helps students formulate knowledge-and scaffold-specific questions and receive targeted help during problem solving.

This project investigates whether large language models (LLMs) can diagnose students’ cognitive skills from handwritten math responses. By introducing the MathCog benchmark and evaluating 16 LLMs, we reveal that even advanced models struggle to infer students' thinking process from fragmented visual and textual cues.