HITS Inc.
AI Researcher
Oct. 2026 – Present
Seonghwan Kim
AI Researcher, HITS Inc.
Machine learning for atomistic simulation & drug discovery
About
Hi! I am an AI Researcher at HITS Inc., an AI-driven drug discovery company. Before joining HITS, I was an AI-CRED Research Fellow at KAIST, working with Prof. Woo Youn Kim and Prof. Sungsoo Ahn. I received my Ph.D. in Chemistry from KAIST, where I worked in the Intelligent Chemistry Lab under the supervision of Prof. Woo Youn Kim.
My research focuses on developing machine learning methods grounded in physical and chemical principles to accelerate atomistic simulations. I have built generative models for molecular structures using diffusion processes, flow matching, and Schrödinger bridges—with a particular emphasis on navigating atomistic energy landscapes, including transition state prediction and molecular structure optimization with quantum-chemical accuracy.
Building on this foundation, I am expanding toward simulation-driven approaches for drug discovery, including molecular dynamics acceleration, free energy calculations, and enhanced sampling of rare events. My goal is to make previously intractable atomistic computations feasible by combining rigorous mathematical frameworks with scalable GPU-based infrastructure.
News
- 2026.10Joined HITS Inc. as an AI Researcher.
- 2026AdaPert accepted to ICML 2026.
- 2026Riemannian Denoising Model published in Nature Computational Science.
- 2026FragFM accepted to ICLR 2026.
- 2025.08Received Ph.D. in Chemistry from KAIST; started as AI-CRED Research Fellow.
Experience
Korea Advanced Institute of Science and Technology (KAIST)
AI-CRED Research Fellow · with Prof. Woo Youn Kim and Prof. Sungsoo Ahn
Aug. 2025 – Oct. 2026
Education
Korea Advanced Institute of Science and Technology (KAIST)
Ph.D. in Chemistry · Advisor: Prof. Woo Youn Kim
Sep. 2019 – Aug. 2025
Korea Advanced Institute of Science and Technology (KAIST)
B.S. in Mathematics and Chemical & Biomolecular Engineering (Double Major)
Mar. 2013 – Feb. 2019
Publications * equal contribution · † corresponding
Nat. Comput. Sci. 2026
Riemannian Denoising Model for Molecular Structure Optimization with Chemical Accuracy
Nat. Comp. Sci. 6, 134–144, 2026
ICML 2026
Learning Adaptive Perturbation-Conditioned Contexts for Robust Transcriptional Response Prediction
ICLR 2026
FragFM: Hierarchical Framework for Efficient Molecule Generation via Fragment-Level Discrete Flow Matching
NeurIPS 2025 AI4D3
MoAgent: A Hypothesis-Driven Multi-Agent Framework for Drug Mechanism of Action Discovery
NeurIPS 2024
GeoTMI: Predicting Quantum Chemical Property with Easy-to-Obtain Geometry via Positional Denoising
Nat. Commun. 2024
Diffusion-based generative AI for exploring transition states from 2D molecular graphs
Nat. Commun. 15, 341, 2024
JCTC 2023
Dynamic Precision Approach for Accelerating Large-Scale Eigenvalue Solvers in Electronic Structure Calculations on Graphics Processing Units
J. Chem. Theory Comput. 2023, 19, 5, 1457–1465
Awards
- Grand Prize, Samsung Electronics DS Division Industry-Academia Cooperation2024
- 2nd Place, Samsung AI Challenge2023
- Recipient, 9th EDISON Computational Chemistry Software Competition2019
Academic Service
- Journal Reviewer Nat. Comput. Sci. (2025), PNAS (2024), Nat. Mach. Intell. (2024), IEEE (2025)
- Conference Reviewer NeurIPS (2024, 2025), ICLR (2025, 2026)