About

I am a postdoctoral fellow at the Innocore AI-CRED Institute, KAIST, advised by Prof. Sungsoo Ahn and Prof. Woo Youn Kim. I received my Ph.D. in Computer Science from Seoul National University under Prof. Sun Kim, where my dissertation on subgraph-informed hierarchical learning was recognized with the Outstanding Doctoral Dissertation Award. My research develops graph-based and multi-modal AI for drug discovery and biomedical reasoning, integrating molecular, cellular, sequential, textual, and knowledge-graph signals across the preclinical pipeline, with a growing focus on LLM-based scientific reasoning over biological knowledge toward trustworthy, biology-grounded systems for drug discovery and personalized medicine.

Research

Multi-modal integration for drug discovery

Cross-modal molecule–language models (EMNLP), knowledge-aligned drug-likeness (ICML), and interpretable drug response prediction from cell-line expression.

4 papers →

Robust & generalizable graph learning

Hierarchical semantic environments for OOD graphs (CVPR) and multi-level hypergraphs for clinical notes and documents (ACL Oral, AAAI).

3 papers →

LLM agents for biological reasoning

Perturbation-conditioned contexts for transcriptional response prediction (ICML), query-adaptive path-finding for biomedical KGQA (EMNLP), and progressive multi-agent reasoning over biological knowledge.

3 papers →Current

News

  • 2026.09TriLemma was accepted at NeurIPS 2026.
  • 2026.08AdaPath was accepted at EMNLP 2026 (Main Conference).
  • 2026.05AdaPert was accepted at ICML 2026.
  • 2025.08MV-CLAM was accepted at EMNLP 2025 (Findings).
  • 2025.08I started my postdoctoral fellowship at the AI-CRED Institute, KAIST.
Earlier
  • 2025.05Boundre.E was accepted at ICML 2025.
  • 2025.04Two papers were accepted at the MLGenX Workshop at ICLR 2025.
  • 2025.03One paper was accepted at IEEE TCBB.
  • 2025.02I received the Outstanding Doctoral Dissertation Award from the Department of CSE, SNU.
  • 2024.12I defended my Ph.D. dissertation, "Subgraph-informed Hierarchical Learning in Clinical and Biomedical Domains."
  • 2024.09I received the Youlchon AI Star Fellowship.
  • 2024.04One paper was accepted at Computational and Structural Biotechnology Journal.
  • 2024.02One paper was accepted at CVPR 2024.
  • 2024.02One paper was accepted at the ICLR 2024 Tiny Papers track.
  • 2024.02I received the 30th Samsung HumanTech Paper Award.

Publications

* Equal contribution

Honors & Awards

  • 2025Outstanding Doctoral Dissertation Award, CSE, Seoul National University
  • 2024Youlchon AI Star Fellowship
  • 202430th Samsung HumanTech Paper Award
  • 202227th Samsung HumanTech Paper Award, Bronze ↗
  • 2022AIIS Retreat Best Poster Award ↗
  • 2022AAAI Student Scholar Scholarship
  • 2020–22SNU Global Scholarships II

Education

  • 2020–24Ph.D. in Computer Science, Seoul National UniversityAdvisor: Prof. Sun Kim · Thesis: Subgraph-informed Hierarchical Learning in Clinical and Biomedical Domains
  • 2018–20M.S. in Computer Science, Seoul National University
  • 2014–18B.S. in Software Engineering, Jilin University

Multi-modal integration for drug discovery

Most biomedical questions only become answerable when several modalities are read together. I build models that couple molecular structure with text, gene expression, and biological knowledge graphs.

MV-CLAM: Multi-View Molecular Interpretation with Cross-Modal ProjectionEMNLP '25
Data · Molecular graph, SMILES, text descriptions
Predicting Drug-likeness via Biomedical Knowledge AlignmentICML '25
Data · Molecular structure, biomedical knowledge graph, approved drugs
Improving Reaction Yield Prediction with Chemical Atom-Level Reaction LearningJCIM '25
Data · Atom-mapped reactions, yield records
DRPreter: Interpretable Anticancer Drug Response PredictionIJMS '22
Data · Cell-line gene expression, drug graphs, pathway knowledge graph

Robust & generalizable graph learning

A model is only useful if it holds up on data it was never trained on. I design graph methods that stay reliable under distribution shift and transfer across domains.

Improving OOD Generalization in Graphs via Hierarchical Semantic EnvironmentsCVPR '24
Method · Invariant learning with hierarchical environments, on OOD molecular property splits
Clinical Note Owns its Hierarchy: Multi-Level Hypergraph Neural NetworksACL '23
Method · Multi-level hypergraph over clinical notes (EHR)
Sparse Structure Learning via GNNs for Inductive Document ClassificationAAAI '22
Method · Sparse structure learning on document graphs

LLM agents for biological reasoning

My current focus: agents that reason over biological knowledge to model cells and predict how they respond to perturbations.

AdaPath: Query-Adaptive Path-Finding via Path-Bank for Multi-Hop Implicit Biomedical KGQAEMNLP '26
Method · Query-adaptive path-finding over a biomedical knowledge graph for multi-hop question answering
Learning Adaptive Perturbation-Conditioned Contexts for Robust Transcriptional Response PredictionICML '26
Data · Perturbation transcriptomics, gene descriptions, biological knowledge graph
Progressive Multi-Agent Reasoning for Biological Perturbation PredictionICMLw '26
Method · Progressive multi-agent LLM reasoning over biological knowledge