Jeonghun Cho

Jeonghun (Jay) Cho

E-mail CV Github LinkedIn Google_Scholar Semantic_Scholar

About

I am a Ph.D. student in Computer Science at University of Maryland, advised by Prof. Heng Huang. My research focuses on post-training and inference-time scaling for large reasoning models, including reinforcement learning with verifiable rewards (RLVR), policy optimization, value estimation, and preference optimization.

Prior to joining UMD, I earned my M.Sc. degree in Artificial Intelligence at POSTECH, where I worked with Prof. Gary Geunbae Lee in the Natural Language Processing Group. Before that, I received a B.Sc. degree in Electrical Engineering at Hanyang University ERICA.

News

Publications

Self-Correcting Code Generation Using Small Language Models
Jeonghun Cho, Deokhyung Kang, Hyounghun Kim, and Gary Geunbae Lee
Findings of EMNLP 2025. [paper] [code]
Retrieval-Augmented Fine-Tuning With Preference Optimization For Visual Program Generation
Deokhyung Kang*, Jeonghun Cho*, Yejin Jeon, Sunbin Jang, Minsub Lee, Jawoon Cho, and Gary Geunbae Lee
Proceedings of ACL 2025. [paper] [press]
K-COMP: Retrieval-Augmented Medical Domain Question Answering With Knowledge-Injected Compressor
Jeonghun Cho and Gary Geunbae Lee
Proceedings of NAACL 2025. [paper] [code]
Medical Domain Retrieval-Augmented Question Answering through Entity-based Context Extraction
Jeonghun Cho and Gary Geunbae Lee
Proceedings of KCC 2024. (Best Paper, Top 1.7%) [paper] [code]

Education

Ph.D., Computer Science, University of Maryland, 2026 -

M.Sc., Artificial Intelligence, Pohang University of Science and Technology, 2023 - 2025

B.Sc., Electrical Engineering, Hanyang University ERICA, 2018 - 2023