Published numerous papers in international conferences like NeurIPS, ICDE, CIKM, and ACL, covering a wide range of topics from mitigating overthinking in tool reasoning to creating a personalized visual persuasion image dataset.
Research Experience
Serves as an Assistant Professor focusing on Conversational AI, Multimodal Generative Diffusion Models, and the Inner Workings of Language Models: Interpretation and Intervention. Teaches courses such as 'Conversational AI' and 'Multimodal Generative Diffusion Models'.
Background
Assistant Professor at the Graduate School of Data Science, Seoul National University. Research interests include but are not limited to: agents grounded in physical devices, software tools, and multimodal data; reasoning enhancement via RL and multi-agent simulation; knowledge representation and editing for language models; and pluralistic value alignment and steering for language models.
Miscellany
Encourages students interested in interning at his lab to take his 'Conversational AI' course.