Hello! 안녕하세요! I am Jaeung Lee, an AI Research Engineer at Samsung Research.

My research interests lie in visualization for machine learning, with a particular focus on how interactive interfaces can support model evaluation and facilitate human-centered decision-making.

Before joining Samsung Research, I received my M.S. and B.S. degrees in Computer Science and Engineering from Sungkyunkwan University in 2026 and 2024, respectively. During my master’s program, I worked at the Interactive Data Computing Lab under the supervision of Prof. Jaemin Jo.

News

Last updated 2026.08.31
  1. Aug 2026

    🎉 Unlearning Depth Score was accepted to EMNLP 2026 (Main) with a top 4.1% score.

  2. Jul 2026

    👾 New personal website launched!

  3. Jun 2026

    ⭐ Unlearning Comparator reached 100 GitHub stars!

  4. Mar 2026

    🚀 Joined Samsung Research as an AI Research Engineer.

Experience

  1. Samsung Electronics

    Research Engineer @ Samsung Research

    Seoul, Korea

    Mar 2026 —Present

    AI System Team, AI Center

    • Designing next-generation on-device architectures for vision-language and embedding models.

  2. Sungkyunkwan University

    Research Assistant @ IDC Lab

    Suwon, Korea

    Sep 2024 —Feb 2026

    Research

    • Visualization and Human-Computer (AI) Interaction for ML, AI safety, and mechanistic interpretability.
    • Authored 5 papers: 2 first-authored (TVCG, EMNLP), 2 co-authored (VIS, WWW), and 1 in progress.
    • Built PediaViz, a clinical visualization system for pediatric chronic care, with NAVER and SNU Hospital.
    • Released pytorch-cka, an open-source PyTorch library for layer-wise model similarity (CKA) with 7K+ downloads.

    Teaching Assistant

    • Fundamentals of Machine Learning (2024 Fall, 2025 Spring, 2025 Fall). See course materials
    • Ranked 1st in student evaluation scores for three consecutive semesters.

  3. deep daiv.

    Content Reviewer & Project Mentor

    Feb 2024 —Jan 2025

    Content Review

    • Reviewed AI articles and newsletters before publication for deep daiv., the largest Korean student AI community on Instagram with 40K+ followers.

    Project Mentoring

    • Mentored explainable AI (XAI) and computer vision projects for 100+ members.

  4. Sungkyunkwan University

    Undergraduate Researcher

    Suwon, Korea

    Jul 2023 —Aug 2024

    IDC Lab Jan 2024 — Aug 2024

    • Contributed to Bavisitter, a system that folds visualization design guidelines into LLM-based chart authoring.

    VSCLab, now V-Lab @ Yonsei University Jul 2023 — Dec 2023

Education

  1. Sungkyunkwan University

    M.S. in Computer Science and Engineering

    Suwon, Korea

    Sep 2024 —Feb 2026

    GPA: 4.43/4.5 · Graduated early

  2. Sungkyunkwan University

    B.S. in Computer Science and Engineering

    Suwon, Korea

    Mar 2018 —Aug 2024

    Overall GPA: 4.32/4.5 · Major GPA: 4.4/4.5

    • Summa Cum Laude, Ranked 1st in the cohort
    • Military service in the Republic of Korea Army (July 2020–January 2022)

Publications

  1. DECODE: Diagnosing Interpretability Failures in Automated Interpretation Process

    Ongoing work.

  2. Measuring the Depth of LLM Unlearning via Activation Patching

    Jaeung Lee, Dohyun Kim, and Jaemin Jo

    Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP, 2,719 / 17,669 = 15.4%), Budapest, Hungary, 2026.

    ProjectGitHubPaper

  3. Suppression or Deletion: A Restoration-Based Representation-Level Analysis of Machine Unlearning

    Yurim Jang, Jaeung Lee, Dohyun Kim, Jaemin Jo, and Simon S. Woo

    Proceedings of the ACM Web Conference 2026 (WWW Short, 123 / 525 = 23.4%), Dubai, United Arab Emirates, 2026.

    GitHubPaper

  4. Unlearning Comparator: A Visual Analytics System for Comparative Evaluation of Machine Unlearning Methods

    Jaeung Lee, Suhyeon Yu, Yurim Jang, Simon S. Woo, and Jaemin Jo

    IEEE Transactions on Visualization and Computer Graphics, vol. 32, no. 3, pp. 2852–2867, 2026.

    ProjectVideoGitHubPaper

  5. Bavisitter: Integrating Design Guidelines into Large Language Models for Visualization Authoring

    Jiwon Choi, Jaeung Lee, and Jaemin Jo

    Proceedings of the IEEE Visualization & Visual Analytics (IEEE VIS), St. Pete Beach, USA, 2024.

    GitHubPaper

More

Honors & Awards

  1. Aug 2024

    DX Scholarship, Samsung Electronics

    Full tuition for the entire master’s program (2024–2025).

  2. Aug 2024

    Graduate Fellowship, Sungkyunkwan University

    Ranked 1st in Computer Science and Engineering

    • Full tuition for the entire master’s program (2024–2025)
  3. Aug 2023

    Best Award, AI Challenge, Ministry of Science and ICT

    Developed a semantic segmentation model to detect buildings in satellite imagery.

    • Ranked 2nd among approximately 1,000 participants
    • Received the Institute of Information & Communications Technology Planning & Evaluation (IITP) Director’s Award
  4. Aug 2023

    Merit Scholarship, Sungkyunkwan University

    Awarded to the top 3% of Computer Science and Engineering

    • Covered 70% of tuition (2023 Spring, 2023 Fall, 2024 Spring)
  5. Feb 2018

    Sungkyun Software Scholarship, Sungkyunkwan University

    Full tuition scholarship (2018–2019).

Skills

  1. Research

    Visual Analytics, Mechanistic Interpretability, NLP, AI Safety

  2. AI / ML

    PyTorch, Hugging Face, scikit-learn, Matplotlib

  3. Frontend

    React, TypeScript, Visualization (D3, Vega, …)

  4. Backend

    Python, FastAPI, Docker, PostgreSQL, AWS (S3, EC2, …)

  5. Design

    Figma, Design Study, User Research, Prototyping

  6. Certificate

    Samsung S/W Certificate (PRO)