Kyungmin Min 민경민

About

I am a Ph.D student at Seoul National University, advised by Prof. Kyomin Jung. My research focuses on building reliable and trustworthy multimodal AI systems. I study how large vision-language models perceive, reason, and sometimes hallucinate — and develop methods to make their outputs more faithful and grounded. My recent work spans hallucination mitigation in LVLMs, evaluation methodology for multimodal models, and experience-driven memory systems for AI agents. I am particularly interested in enabling agentic systems that learn from their own experience and improve over time without additional training, and in understanding why vision-language models lose image grounding as sequences grow longer.