Seoul, South Korea

Hi!

I'm Namho

GitHub
Arrow icon
LinkedIn
Arrow icon
Google Scholar
Arrow icon
Resume
Arrow icon
📚 Currently a [2nd yr] Ph.D. candidate @AIPR Lab under the supervision of Professor Kee-Eung Kim @KAIST AI

My current focus is on the applications of Reinforcement Learning and Robotics within Agriculture.
‍
🏋🏼 USAPL Korean Nationals '24 placing 🥉and at the Collegiate Nationals '24 placing 18th [-110kg]

About

My passion lies in conducting rigorous and salient research in developing tangible software applications to address true user needs and issues. I thoroughly enjoy tackling complex problems through an interdisciplinary manner whilst evaluating the efficacy in a scientific approach.

During my free time, I enjoy power/weight-lifting and  Brazilian jiu jitsu.

Teaching

[KAIST] AI611 (Deep Reinforcement Learning) F26
[KAIST] AI503 (Mathematics for AI) F25
[UMich] EECS 486 (Information Retrieval and Web Search) W23 & W24: lecture
[UMIch] EECS 587 (Parallel Computing) F23: lecture

Selected work and contributions

NVIDIA CES 2026 RTX and ACE announcement
NVIDIA • ACE SDK

NVIDIA ACE

Contributed to research and development, with work showcased in NVIDIA's CES 2026 announcement (Total War: PHARAOH AI advisor). Announcement

Publications

EMNLP 2026

Main Conference

Dynamic Context Allocation for Concurrent Multi-Agent LLMs

J. Na, D. E. Matsunaga, N. H. Koh, J. Lee, P. Poupart, K. Kim

EMNLP 2026

Findings

Same Tasks, Different Apps: Why Mobile GUI Agents Fail to Generalize?

T. Tran*, N. H. Koh*, A. Jain, D. E. Matsunaga, K. Kim

* Equal contribution

Under review

Three manuscripts in double-blind review

1 submission to WACV 2027 · 2 submissions to AAAI 2027