Daejeon, Korea ●Hi, I’m Donggyu
AI for economic
and social questions.
PhD student in Data Science at KAIST, studying how AI systems reason about economic and social phenomena.
I work across large language models, satellite imagery, and computational social science—especially when reliable data are scarce.
Selected work
Research
NeurIPS 2026 EconML Workshop
EconCausal: A Context-Aware Causal Reasoning Benchmark for Large Language Models in Social Science
NeurIPS 2026 SocialAgent Workshop · NeurIPS 2026 UserSim Workshop
From Simulated Citizens to Simulated Deliberation: Challenges in Representation and Interaction
ACM GOODIT 2025
LLM-Driven Socio-Economic Estimation for Visegrád Countries
Places I’ve learned from
Experience
Hover for a quick hint. Click a place to see what I worked on.
What I worked on
Adapted and enhanced satellite-based crop classification methods for rice, maize, and other major crops, improving transferability and performance in data-scarce regions. The work was conducted with Prof. David Lobell and Prof. Jennifer Burney, in collaboration with the Ministry of Unification of the Republic of Korea.
What I worked on
Estimated socioeconomic indicators using satellite imagery and large language and vision-language models, with applications spanning North Korea, other developing regions, and the Visegrád Group. Conducted in the Data Science for Humanity Group, advised by Prof. Meeyoung Cha.
What I worked on
Built a synthetic-data pipeline with Stable Diffusion and DreamBooth to generate smartphone defect images for Galaxy quality-inspection models and improve adaptation to new product launches. The project was conducted in Samsung Electronics Global Research’s Smart Factory Group.
🏆 Best Internship Project AwardWhat I worked on
Studied diffusion-based generative models through foundational work including DDPM, DDIM, and Stable Diffusion, then presented their methods and key ideas to the research group.
What I worked on
Developed a prototype recommender system for a Korean financial company to recommend insurance riders from customer and product data, exploring collaborative filtering and matrix-factorization approaches such as SVD.
Details follow my current LinkedIn profile. KAIST lab names use the labels from this website. View profile ↗
Education
KAIST
Ph.D. Data Science · Sep 2026 – Present
M.S. Data Science · Sep 2024 – Aug 2026
B.S. Computer Science · Mar 2017 – Aug 2024
Economics minor
Exchange at EPFL · Aug 2022 – Feb 2023
Latest note
01. I started researching investments on my own
Small capital, simple rules, and the hope of building a research process I could run alone.
Read writing ↗