Minki Hong
Hi! Iβm Minki Hong, an M.S. student in Computer Science and Artificial Intelligence at Dongguk University, advised by Professor Jihie Kim. I received my B.S. degree in Computer Science & Engineering from Dongguk University.
My research interests lie in developing AI systems that align with human social norms and values. I focus on the intersection of generative AI, social norms, and bias mitigation, with a particular emphasis on fairness and inclusivity. I am especially fascinated by how cultural and social contexts shape communication and how AI models can better capture these nuances. In addition, I explore methods to improve the quality of generative AI responses in low-resource environments, aiming to make AI more accessible and equitable across diverse linguistic and cultural settings.
I am also deeply interested in eliciting reasoning capabilities in both large language models (LLMs) and visionβlanguage models (VLMs). I find it exciting that different prompting and training strategies can enable smaller models to approach the performance of much larger ones, and I am actively exploring methods to bridge this gap.
Currently, I am spending six months as a visiting researcher at the Robotics Institute of Carnegie Mellon University under the supervision of Professor Jean Oh. During this period, I am studying the interaction between robotics and generative AI, investigating how these two domains can complement and enrich each other.
If you would like to connect or chat, please feel free to reach out:
π§ jackyh1@dgu.ac.kr | π§ minkih@andrew.cmu.edu | π LinkedIn | π My blog
π° News
- Jan 04, 2026 β EACL 2026 (Findings) paper accepted! π
- Dec 12, 2025 β IASEAI 2026 paper accepted! π
- Nov 11, 2025 β π SAC Highlights at EMNLP 2025 π
- Sep 15, 2025 β Selected for oral presentation at EMNLP 2025! π
- Aug 19, 2025 β EMNLP 2025 paper accepted! π
π Publications
- Under Review β Evaluating Demographic Misrepresentation in Image-to-Image Portrait Editing
- EACL 2026 (Findings) β VisDoT : Enhancing Visual Reasoning through Human-Like Interpretation Grounding and Decomposition of Thought
- Mathematics β Systematic Integration of Attention Modules into CNNs for Accurate and Generalizable Medical Image Diagnosis
- IASEAI 2026 β Exposing Blindspots: Cultural Bias Evaluation in Generative Image Models
- EMNLP 2025 β NormGenesis: Multicultural Dialogue Generation via Exemplar-Guided Social Norm Modeling and Violation Recovery
π Education
- M.S. in Computer Science and Artificial Intelligence. Dongguk University. Mar. 2024 - Present
- B.S. in Computer Science & Engineering. Dongguk University. Mar. 2018 - Feb. 2024
