M.S. Student, Industrial Engineering @ Seoul National University DSBA Lab (Data Science & Business Analytics Lab), advised by Prof. Pilsung Kang
π Homepage Β· βοΈ hankyeol@snu.ac.kr Β· πΌ LinkedIn Β· π Google Scholar
I am an M.S. student in Industrial Engineering at Seoul National University (SNU), and I received my B.S. in Physics and Astronomy from SNU. My past work spanned physics-informed neural networks and LLM confidence; I am now working on spatiotemporal diffusion and visual reasoning of VLMs. In the long term, I am pursuing physical AI.
PAST |
Physics-Informed Neural Networks Β· LLM Confidence |
NOW |
Spatiotemporal Diffusion Β· Visual Reasoning of VLMs |
NEXT |
Physical AI |
-
Asking Is Not Enough: Protocol Sensitivity in LLM Confidence Calibration Hankyeol Kim, Pilsung Kangβ Under review, 2026 β [Paper] [Code]
-
naPINN: Noise-Adaptive Physics-Informed Neural Networks for Recovering Physics from Corrupted Measurement Hankyeol Kim, Pilsung Kangβ Under review, 2026 β [Paper] [Code]
- Seoul National University β M.S. in Industrial Engineering, DSBA Lab (Mar 2025 β Present)
- Seoul National University β B.S. in Physics and Astronomy (Mar 2020 β Feb 2025)
- Lunosoft Inc. Β· Project Manager (Feb 2024 β Nov 2024) Druid Rising: Idle RPG Β· 2026 Trash King: Idle Tycoon Β· Love Revolution: Hidden Object
- AIARA Inc. Β· Project Management Team Leader / Project Manager / App Developer (Dec 2022 β Feb 2024) Projects for Samsung Electronics, E-MART, Megastudy Education, Seoul National University Hospital, and LUXROBO
- SBA SeSAC Β· Boost Camp Mentor, Service Design (Dec 2023 β Mar 2024)
- SK hynix Inc. β Machine Learning Model for CapL Prediction via InLine ICBV Learning (i-TAP) (May 2026 β Present)
- Meritz Fire & Marine Insurance β Contextual BanditβBased Policy Optimization Framework for Insurance Restructuring Strategy Selection and Combination (Apr 2026 β Present)
- Samsung Fire & Marine Insurance β Data Construction and Optimization for Training/Inference of an Information Retriever (Mar 2025 β Mar 2026)
- ETRI β Anomalous Situation Detection Based on Individual Daily Patterns Using Location Information (Jul 2025 β Nov 2025)