Hojae Lee 이호재
[CV], [Linkedin], [Google Scholar], [GitHub], [hojae@arizona.edu]
Embedded Files
Hi👋. I'm Hojae, a PhD student at the University of Arizona, advised by Prof. Mihai Surdeanu. My primary research lies in the area of Natural Language Processing; I work on how human behaviors/virtues can be computationally applied to language models, so that they behave like us and understand our world more appropriately. I am currently interested in (but not limited to) the research topics below:
Complex Reasoning: Improving reasoning abilities (which is a core ability of human intelligence) of language models through methods such as multi-step reasoning and information retrieval systems.
Efficient NLP: Devising methods to decrease training/inference costs of language models while preserving their performances, such as knowledge distillation.
Computational Linguistics: Investigating natural language understanding of language models through the lens of linguistics, such as construction grammar.
News
News
[2025/04] Excited to share that I am accepted to the University of Arizona's CS PhD program which will begin in Fall 2025 🌵. I have officially accepted the offer!!
[2024/10] I joined the Language Technology Research Institute (LTRI) at Hankuk University of Foreign Studies as a visiting predoctoral researcher.
[2024/09] Our paper "Mentor-KD: Making Small Language Models Better Multi-step Reasoners" has been accepted to EMNLP 2024. See you all in Miami ⛱️!!
[2024/05] I successfully defended my Master's thesis🧑🎓!
Education
Ph.D. in Computer Science (In Progress)
University of Arizona @ Tucson, AZ, USA ― Aug 2025 to Present
Advisor: Prof. Mihai Surdeanu
M.S. in Computer Science and Engineering
Korea University @ Seoul, South Korea ― Mar 2022 to Aug 2024
Advisor: Prof. SangKeun Lee
B.A. in English Linguistics, B.L.S. in Language & AI (double major)
Hankuk University of Foreign Studies @ Seoul, South Korea ― Mar 2016 to Feb 2022
Publications
*: equal contribution
Mentor-KD: Making Small Language Models Better Multi-step Reasoners
Hojae Lee*, Junho Kim*, SangKeun Lee
EMNLP 2024 (long)
Improving Chain-of-Thought Distillation with Annotator Models
Hojae Lee
KCC 2024
Other Interests
I do music for fun, and enjoy re-arranging/orchestrating musical pieces as well as playing the piano/synthesizer.
Google Sites
Report abuse