I build foundational and applied methods for AI measurement and the learning sciences.

I focus on the intersection of AI × Education × Psychometrics to achieve the simple, yet difficult, goal of education: helping students prepare for emerging opportunities in a world shaped by research, technology, and AI.

About

01 / 03
Kaz Fukuhara

I'm an MPhil student at the UCL Knowledge Lab, University College London, working with Professor Mutlu Cukurova. Previously, I completed my M.S. at the Stanford Graduate School of Education, working with Associate Professor Benjamin Domingue.

I study the potential of large language models (LLMs) for student simulation combined with psychometrics, spanning applications in test design, teacher training, and personalized learning. I extend this work to measuring and developing teachers’ competency in AI, finding ways to help teachers navigate their experiences with rapidly advancing technologies.

Now
UCL Knowledge Lab, MPhil — advisor: M. Cukurova
Previously
Stanford GSE, M.S. — advisor: B. Domingue
Interests
AI measurement · IRT / CDM · LLM-simulated learners · belonging-centered instruction
Based in
London, United Kingdom

Research

02 / 03
2026
AIME-Con 2026

Can LLMs "Interpret" Psychometric Parameters? Evaluating "Over-Knowledge Problem" in LLM Student Simulation

Fukuhara, Domingue

2026
Findings of EMNLP 2026

A Dataset for Modeling Iterative Problem-Solving

Patel, Truong, Nguyen, Fukuhara, Haber, Koyejo, Domingue

2025
Working Paper

Evaluating LLM Adaptation Techniques to Support Belonging-Centered Instruction in Elementary Mathematics Classrooms

Shi, Fukuhara

Projects

03 / 03
In Build · Stanford
Platform / Ecosystem

ResearchC

A research-industry ecosystem for university students.

I work on improving the existing research ecosystem in universities by building ResearchC, a Stanford-based platform that helps student researchers find collaborators, mentorship, and cross-departmental research opportunities.

Through ResearchC, I envision higher-education students leveraging research for their own goals — in academia or in industry — to maximise their learning opportunities.

In Build · Tokyo
Platform / News App

EduUpdate

A news app for early-career educators.

I am building EduUpdate, a personalized news and learning app designed to help aspiring and early-career educators stay informed about the latest developments in education.

From curated news feeds and education policy updates to foundational theories and digital transformation practices, EduUpdate makes essential knowledge accessible and actionable for the next generation of educators.