Sungjun Lim

Trustworthy AI,Bayesian Deep Learning,Uncertainty Quantification,Mechanistic Explainability,Distribution Shift Robustness,Probabilistic Representations

About Me

Name: Sungjun, Lim

Birth: Sep. 11, 1998

Department: Statistics and Data Science, Yonsei University

Email: lsj9862@gmail.com

Hello! I am Sungjun Lim, a graduate researcher at Yonsei University's Statistics and Data Science department.

My research centers on building robust and trustworthy AI systems, focusing on uncertainty quantification, robustness under distribution shift, and mechanistic interpretability to understand model behavior in out-of-distribution settings.

Currently, I am working on uncertainty-aware and geometry-aware interpretability methods for large-scale neural networks, aiming to improve the robustness and faithfulness of explanations in real-world scenarios.

Career

B.S.

Statistics, University of Seoul

Mar. 2017 – Aug. 2022

Undergraduate Research Assistant

MLAI Lab, University of Seoul

Jun. 2021 – Aug. 2022

M.S.

Artificial Intelligence, University of Seoul

Advisor: Kyungwoo Song

Sep. 2022 – Feb. 2024

Ph.D. Student

Statistics and Data Science, Yonsei University

Advisor: Kyungwoo Song

Mar. 2024 – Present

Visiting Researcher

Computer Science, Australian National University

Advisor: Lexing Xie

Jul. 2025 – Aug. 2025

Publications

Google Scholar

📘 Peer Reviewed

  1. Language model-guided student performance prediction with multimodal auxiliary information
    • Changdae Oh, Minhoi Park, Sungjun Lim, Kyungwoo Song
    • Expert Systems with Applications (ESWA) 2024

  2. GFML: Gravity Function for Metric Learning
    • Hoyoon Byun, Sungjun Lim, Kyungwoo Song
    • Engineering Applications of Artificial Intelligence (EAAI) 2025

  3. Robust Optimization for PPG-based Blood Pressure Estimation
    • Sungjun Lim, Taero Kim, Hyeonjeong Lee, Yewon Kim, Minhoi Park, Kwang-Yong Kim, Minseong Kim, Kyu Hyung Kim, Jiyoung Jung, Kyungwoo Song
    • Biomedical Signal Processing and Control (BSPC) 2025

  4. Brain-inspired Lp-Convolution benefits large kernels and aligns better with visual cortex
    • Jae Kwon, Sungjun Lim, Kyungwoo Song, C. Justin Lee
    • International Conference on Learning Representations (ICLR) 2025

  5. Sufficient Invariant Learning for Distribution Shift
    • Taero Kim, Subeen Park, Sungjun Lim, Yonghan Jung, Krikamol Muandet, Kyungwoo Song
    • Computer Vision and Pattern Recognition (CVPR) 2025

  6. Flat Posterior Does Matter For Bayesian Model Averaging
    • Sungjun Lim, Jeyoon Yeom, Sooyon Kim, Hoyoon Byun, Jinho Kang, Yohan Jung, Jiyoung Jung, Kyungwoo Song
    • Uncertainty in Artificial Intelligence (UAI) 2025

  7. Uncertainty Aware Contrastive Decoding
    • Hakyung Lee, Subeen Park, Joowang Kim, Sungjun Lim, Kyungwoo Song
    • Association for Computational Linguistics (ACL) 2025 Findings

  8. COVID-19 Prediction with Doubly Multi-task Gaussian Process
    • Sooyon Kim, Yongtaek Lim, Sungjun Lim, Gyeongdeok Seo, Jihee Kim, Hojun Park, Jeahun Jung, Kyungwoo Song
    • Journal of Biomedical Informatics 2025

  9. Causal Effect Variational Transformer for Public Health Measures and COVID-19 Infection Cluster Analysis
    • Jinho Kang, Sungjun Lim, Kyungwoo Song
    • Conference on Information and Knowledge Management (CIKM) 2025

  10. Data Adaptive Stochastic Ensemble Net: Optimizing Infection Predictions for COVID-19 Cluster Analysis
    • Sungjun Lim, Yongtaek Lim, Hojun Park, Junggu Lee, Jaehun Jung, Kyungwoo Song
    • IEEE Journal of Biomedical and Health Informatics 2025

  11. RAILL : Retrieval-Augment and Instruction Tuning for Low-resource Language Model Training
    • Youngjun Choi, Sungjun Lim, Minhoi Park, Jaekyeong Jung, TaeKyung Kim, Hosik Choi, Kyungwoo Song
    • IEEE Big data 2025 (Short)

  12. Semi-Supervised Preference Optimization with Limited Feedbacks
    • Seonggyun Lee, Sungjun Lim, Seojin Park, Soeun Cheon, Kyungwoo Song
    • International Conference on Learning Representations (ICLR) 2026

  13. Uncertainty-driven Embedding Convolution
    • Sungjun Lim, Kangjun Noh, Youngjun Choi, Heeyoung Lee, Kyungwoo Song
    • International Conference on Learning Representations (ICLR) 2026

📝 Under-Review

  1. DDRL: A Diffusion-Driven Reinforcement Learning Approach for Enhanced TSP Solutions
    • Joowang Kim, Jeyoon Yeom, Gyeongdeok Seo, Sungjun Lim, Jae Ha Kwak, Heejun Ahn, Gyeong-moon Park, Kyungwoo Song

  2. Eigen-Value : Efficient Domain-Robust Data Valuation via eigenvalue-Based Approach
    • Youngjun Choi, Junseong Kang, Sungjun Lim, Kyungwoo Song

  3. PTD: Partial-Teacher Distillation for Efficient MoE Recovery in Large Language Models
    • Hoyoon Byun, Kangjun Noh, SoMin Kim, Heedong Kim, Sungjun Lim, Jaeyoon Shim, Youngjun Choi, Kyungwoo Song

🎓 Workshop

  1. Sufficient Invariant Learning for Distribution Shift
    • Taero Kim, Sungjun Lim, Kyungwoo Song
    • The Sixth Data Science Meets Optimisation (DSO) Workshop at IJCAI 2024

  2. Sequential Treatment Effect Estimation with Variational Transformers: Application to COVID-19 Infection Clusters
    • Jinho Kang, Sungjun Lim, Kyungwoo Song
    • Artificial Intelligence for Time Series Analysis (AI4TS) at IJCAI 2024

  3. Flat Posterior For Bayesian Model Averaging
    • Sungjun Lim, Jeyoon Yeom, Sooyon Kim, Hoyoon Byun, Jinho Kang, Yohan Jung, Jiyoung Jung, Kyungwoo Song
    • Frontiers in Probabilistic Inference Workshop at ICLR 2025

MLAI Projects

MLAI@Yonsei

Explainable AI for Blood Pressure Estimation

  • Funded by ETRI
  • Deal with Uncertainty about Estimation of BP from AI
  • Causality Covid-19

  • Funded by Ministry of Food and Drug Safety
  • Infection prediction based on causal graph
  • Directional GNN for Infection Prediction

  • Funded by Ministry of Food and Drug Safety
  • Infection prediction based on graphical information
  • Infection Prediction Based on Gaussian Process

  • Funded by Ministry of Food and Drug Safety
  • Infection prediction based on robabilistic model
  • Educational Content Relationship Analysis

  • Funded by TIPS
  • Analyze educational content relationship via LLMs and RAG
  • Signal Processing

  • Funded by ADS
  • Develop class incremental algorithm to classify aviation object