I am a PhD student at the KAIST Graduate School of AI under the supervision of Prof. Jaesik Choi. My research focuses on understanding the internal mechanisms of deep learning models, especially how structural organization in vision and multimodal representations supports generalizable reasoning. Recently, I have been particularly interested in the expressive capabilities and failure modes of visual generative models, as well as representation-level interpretability for large-scale vision and vision-language systems. My goal is to better understand how these models generalize in open-ended real-world settings.

Research Interest

  • Generative AI
  • Representation Learning
  • Interpretability
  • Vision-Language Models
  • Computer Vision

Education

  • KAIST, PhD Candidate in Artificial Intelligence, Aug 2022 - Present
  • KAIST, M.S. in Artificial Intelligence, Aug 2020 - Aug 2022
  • Yonsei University, Bachelor's degree in Applied Statistics, Mar 2016 - Aug 2020

Awards

  • Insung Scholarship, 2025
  • KAIST Breakthroughs Spring 2026, selected for KAIST Breakthroughs 50

Selected Publications

Granular Concept Circuits: Toward a Fine-Grained Circuit Discovery for Concept Representations

[paper]

Dahee Kwon, Sehyun Lee and Jaesik Choi

ICCV 2025

Enhancing Creative Generation on Stable Diffusion-based Models

[paper]

Jiyeon Han, Dahee Kwon and Jaesik Choi

CVPR 2025

Understanding Distributed Representations of Concepts in Deep Neural Networks without Supervision

[paper]

Wonjoon Chang, Dahee Kwon and Jaesik Choi

AAAI 2024 (Oral)

See all publications

Talks

  • Understanding Deep Neural Networks Decision-Making Through Exploring Learned Features, AI EXPO KOREA 2024 Workshop
  • Analyzing the Attribute-relevant Featuremaps in Stable Diffusion Models, KCC XAI Workshop
  • Understanding Diffusion-based Generative Models, KAIST XAI Tutorial Series
  • Enhancing Creativity in Text-to-Image Generation, Samsung AI Forum