About
Hi!I'm Hyeonbin.
I’m an incoming Ph.D. student at KAIST STAI Lab, advised by Seongjoon Oh. I previously completed my Master’s at KAIST AI, where I was advised by Minjoon Seo. I received my B.S. in CS from KAIST.
My long-term goal is to build models that reason beyond the human data distribution. To this end, I study the learning signals, data structures, and evaluation methods that enable models to self-train and generalize systematically in multi-step reasoning. Currently, I am interested in iterative RL post-training and agentic systems, particularly autonomous research agents.
I’m always open to collaborations—feel free to reach out!
Experience
-
2024 — Now
KAIST AI
M.S. · Language Model Reasoning
Advisor: Minjoon Seo
-
2025
-
2022
NAVER
Healthcare AI · Internship
EHR summarization
-
2019 — 2024
KAIST SoC
B.S. in Computer Science
GPA 3.96 / 4.3
Reviewer
Conferences
- NeurIPS2026
- ICLR2026
- ICML2026Gold Reviewer (Top 25%)
Journals
- TMLR2026
Publications
Selected Publications
-
Intrinsic Task Symmetry Drives Generalization in Algorithmic Tasks
H Hwang, Y Park ICML 2026
-
Let's Predict Sentence by Sentence
H Hwang, B Jeon, S Kim, J Kim, H Chang, S Yang, S Won, D Lee, Y Ahn, ... COLM 2025 RAM 2 Workshop (Oral)
-
Self-Explore to Avoid the Pit: Improving the Reasoning Capabilities of Language Models
H Hwang, D Kim, S Kim, S Ye, M Seo EMNLP 2024 (Findings)ACL 2024 NLRSE Workshop (Oral)
Others
-
The Coverage Principle: A Framework for Understanding Compositional Generalization
H Chang, J Park, H Cho, S Yang, M Ko, H Hwang, S Won, D Lee, Y Ahn, ... ICLR 2026
-
The CoT Encyclopedia: Analyzing, Predicting, and Controlling how a Reasoning Model will Think
S Lee, S Kim, M Seo, Y Jo, D Go, H Hwang, J Park, X Yue, S Welleck, G Neubig, M Lee, M Seo ICLR 2026
-
Differential Information Distribution: A Bayesian Perspective on Direct Preference Optimization
Y Won, H Lee, H Hwang, M Seo Preprint
-
BiGGen Bench: A Comprehensive Benchmark for Generative Language Models
S Kim, J Suk, ... (others not shown), H Hwang, ... M Seo NAACL 2025 (Best Paper)
-
Knowledge Entropy Decay during Language Model Pretraining Hinders New Knowledge Acquisition
J Kim, H Lee, H Cho, J Jang, H Hwang, S Won, Y Ahn, D Lee, M Seo ICLR 2025 (Oral)AAAI 2025 KnowFM Workshop (Best Paper)
-
FLASK: Fine-grained Language Model Evaluation Based on Alignment Skill Sets
S Ye, D Kim, S Kim, H Hwang, S Kim, Y Jo, J Thorne, J Kim, M Seo ICLR 2024 (Spotlight)