Woosung (Reiss) Koh
Woosung (Reiss) Koh

Visiting Researcher

About Me

👋 Hi! My principal research interest is at the intersection of Language Models and Agentic Behaviour. My past works make [C1] multi-agents more generalizable and significantly more reliable; and [W1] improves agents in noisy environments. I have also made an [C2] agentic LLM-as-a-Judge framework for scalable LM-based (multi-modal) chart generation, with robust improvements across 27B, 70B, and frontier-level LMs. In the past, I worked on [J1] personalization and user-centric machine learning; and [W2] improving time-series representation learning under non-stationarity. [WIP] I am currently working on efficient LM reasoning; but nevertheless interested in a wide range of open problems.

I am perfectly bilingual in English and Korean. I love to chat about research and downstream impact; feel free to reach out to me via email 📧.

Interests
  • Language Model Agents
  • Efficient Models and Learning
  • Reliable and Safe Models
  • Reinforcement Learning
Education
  • Undergraduate

    Yonsei University

  • Diploma Programme

    International Baccalaureate

📚 Publications

*First Author(s), ^Advisor(s)

(2024). C2^2: Scalable Auto-Feedback for LLM-based Chart Generation. NAACL 2025 Main (Oral, Long) (7.7% of Submissions).
(2024). FlickerFusion: Intra-trajectory Domain Generalizing Multi-Agent RL. ICLR 2025 (Prelim. NeurIPS 2024 Open-World Agent Workshop).
(2024). Encoding Temporal Statistical-space Priors via Augmented Representation. IJCAI 2024 STRL Workshop (Oral).
(2024). Curriculum Learning and Imitation Learning for Model-free Control on Financial Time-series. AAAI 2024 AI4TS Workshop (Oral) (Top 27% of Accepted Papers).
(2024). Network-based exploratory data analysis and explainable three-stage deep clustering for financial customer profiling. Engineering Applications of Artificial Intelligence (SCIE, Q1).