Welcome to the Machine and Language Learning Lab (ML3) at Yonsei AI. Our mission is to build AI systems that are both highly capable and aligned with humans. We pursue this mission through machine learning with language, a natural interface between humans and machines, as our central lens while extending our research to other modalities.
Our current research centers on large language models and their multimodal extensions, organized around two complementary directions:
- Human-centered AI: Alignment, personalization, safety, and interpretability
- Generalizable AI: Adaptation, self-improvement, multimodality, and evaluation
Recent News
(show all)| Oct, 2026 | InterPol has received the SAC Highlight Award at EMNLP 2026 (50/17669=0.28%)! |
|---|---|
| Sep, 2026 | Two papers (HuB-J and ReCon) are accepted to NeurIPS 2026! |
| Aug, 2026 | Seven papers (ReFeri, CPR, SyRuP, SPRInG, InterPol, LookBack, and LaRA) are accepted to EMNLP 2026, including 3 Oral Presentations (516/17669=2.92%)! |
| Jul, 2026 | A11yn is accepted to COLM 2026! |
| May, 2026 | RD Clustering for LLM interpretability is accepted to ICML 2026 as Spotlight Presentation (536/23918=2.24%)! |
| Apr, 2026 | Five papers (ReVisiT, EMCee, Critical Dimension, Gap-K, and Reasoning UID) are accepted to ACL 2026, including 2 Oral Presentations (558/12148=4.61%)! |
| Jan, 2026 | Six papers (AMIS, DIRECTER, TITOK, Prior Filtering, RPM, and HEX) are accepted to ICLR 2026! |