Dan (Hojun) Yoo

Dan (Hojun) Yoo

CS PhD student at Purdue
Human-AI Interaction researcher

Purdue University · Computer Science

Human-centered AI agents, built to be understood.

Ph.D. Student in Computer Science at Purdue University

I study how people and AI systems interact, how language models reason, and how we can evaluate intelligent agents in ways that reflect human values.

Advised by Prof. Jason Wu.

I received my B.S. in Computer Science, cum laude, from the University of Notre Dame in May 2026. My earlier research with SaNDwich Lab, MINE Lab, and the FORESEER Group spanned web-agent evaluation, value alignment, reasoning calibration, and benchmark analysis.

Research

Research interests

Human–AI Interaction

Designing systems that help people collaborate with AI agents in useful, usable, and human-centered ways.

Agent Evaluation

Developing rigorous methods that capture interaction quality, behavioral alignment, and full execution processes beyond task success.

LLM Reasoning & Interpretability

Understanding, probing, and calibrating how language models reason and make decisions in real-world systems.

Selected work

Publications

Values and preferences steer web agents along different shopping paths toward different headphone choices.

IUI 2026

The Behavioral Fabric of LLM-Powered GUI Agents: Human Values and Interaction Outcomes

Shows how human values and preferences reshape web-agent reasoning, action trajectories, and final choices.

Simret Araya Gebreegziabher, Yukun Yang, Charles Chiang, Hojun Yoo, Chaoran Chen, Hyo Jin Do, Zahra Ashktorab, Werner Geyer, Diego Gómez-Zará, and Toby Jia-Jun Li

PaperDOI

EvalAgent workflow from configuring agent profiles through trajectory comparison, criteria-based judging, and explanations.

IUI 2026 · Demo

EvalAgent: Interactive Comparative Evaluation of Computer-Using GUI Agents

Makes agent behavior comparable through unified trajectory views, user-defined criteria, and evidence-linked judgments.

Yukun Yang, Simret Araya Gebreegziabher, Hojun Yoo, Charles Chiang, Chaoran Chen, Annalisa Szymanski, Hyo Jin Do, Zahra Ashktorab, Werner Geyer, Diego Gómez-Zará, and Toby Jia-Jun Li

PaperDOI

Updates

News

  1. Started my Ph.D. at Purdue University. I also joined CS 593: Graduate Human-Computer Interaction as a teaching assistant.

  2. Graduated cum laude from the University of Notre Dame with a B.S. in Computer Science.

  3. EvalAgent was published in the IUI 2026 Companion Proceedings Demos Track.

  4. The Behavioral Fabric of LLM-Powered GUI Agents was published at IUI 2026.