Minjae Leemjbooo@kaist.ac.kr / minjae.lee.official@gmail.com |
|
About
I am an AI Algorithm Researcher at FuriosaAI, where I focus on language modeling, inference acceleration, and test-time compute scaling. My general research interest lies in efficiently solving real-world problems using machine learning, with two main threads:
- Agentic and Efficient LLM Inference: Developing methods for efficient inference and AI agents (focusing on test-time compute scaling and reasoning) within language models.
- Applied Healthcare ML: Investigating techniques for encoding and synthesizing Electronic Health Records (EHRs) and facilitating EHR information sharing between institutions.
I received my M.S. from the Kim Jaechul Graduate School of AI (GSAI) at KAIST, advised by Edward Choi, where I wrote my thesis on Exploring Optimal Encoders for Electronic Health Records. I completed my B.S. in Industrial & Systems Engineering at KAIST with Summa Cum Laude honors (GPA 4.0/4.3).
Publications
TABED: Test-Time Adaptive Ensemble Drafting for Robust Speculative Decoding in LVLMs
EACL Findings, 2026 | ICLR Workshop on Scalable Optimization for Efficient and Adaptive Foundation Models, 2026
VersaPRM: Multi-Domain Process Reward Model via Synthetic Reasoning Data
ICML, 2025 🏆 Oral (top ~1%)
Rediscovery of CNN's Versatility for Text-based Encoding of Raw Electronic Health Records
CHIL, 2023 🏆 Oral (top ~10 papers)
Experience
|
AI Algorithm Researcher FuriosaAI, Seoul, Korea Language modeling, inference acceleration, and test-time compute scaling
|
Mar. 2023 – Present |
|
Post-MS Research Assistant KAIST, Seongnam, Korea EHR synthesis and transfer learning in healthcare with MD collaboration
|
Jul. 2023 – Jul. 2024 |
|
Research Intern NAVER CLOVA, Seongnam, Korea Quantization and compression techniques for efficient AI models
|
Dec. 2021 – Feb. 2022 |
Education
|
M.S. in AI Kim Jaechul Graduate School of AI (GSAI), KAIST, Seongnam, Korea Advisor: Edward Choi | Thesis: Exploring Optimal Encoders for Electronic Health Records |
Mar. 2023 |
|
B.S. in Industrial & Systems Engineering; Minor in Economics KAIST, Daejeon, Korea Summa Cum Laude (GPA 4.0/4.3) |
Feb. 2021 |
|
Exchange Program in Informatics (Informatik) Technische Universität Berlin (TUB), Berlin, Germany Mirae Asset Outbound Exchange Student Scholarship |
Apr. – Jul. 2017 |
Invited Talks & Teaching
|
Language Models in Clinical ML: My Journey with Two Questions Center for Advanced Medical Computing and Analysis (CAMCA), Harvard Medical School / MGH |
2025 |
|
How Will DeepSeek-R1 Impact Education's Future? College of Education, Chungnam National University |
2025 |
|
Summer Session for Medical AI (CNN for Chest X-ray classification) Korean Society of Artificial Intelligence in Medicine (KoSAIM), 100+ medical staff and students |
Summer 2023 |
|
AI Short Course Program (Diffusion models) Korean Artificial Intelligence Association (KAIA), 100+ attendees |
Spring 2023 |
|
Machine Learning for Healthcare (AI612) – Teaching Assistant KAIST Graduate Course, 200+ students (Instructor: Edward Choi) |
Spring 2022 |
|
Programming for AI (AI504) – Teaching Assistant KAIST Graduate Course, 200+ students (Instructor: Edward Choi) 🏆 Best Lecture Award (Fall 2021) |
Fall 2021, Fall 2022 |
Awards & Honors
| Summa Cum Laude, KAIST | 2021 |
| National Science and Technology Scholarship, Ministry of Science, Korea | 2021 – 2022 |
| Outbound Exchange Student Scholarship, Mirae Asset | 2017 |
| Academic Excellence Scholarship (2nd place), KAIST | 2016 |
| National Science and Technology Scholarship (merit-based), Ministry of Science, Korea | 2015 – 2016 |
| Undergraduate School Fellowship, KAIST | 2013 – 2016 |
Academic Services
| Reviewer – Machine Learning for Health Symposium (ML4H) | 2023 – 2025 |
| Reviewer – Machine Learning for Healthcare (MLHC) | 2025 |