Sungmin Kang

Hi, I am a postdoctoral researcher at KAIST.

• email: sungmin dot kang at kaist dot ac dot kr
• 🎓 Google Scholar
• 🏠 Homepage

Research Interests

I am interested in building automated tools that help developers do their work.

Education

• Ph.D., Computer Science, KAIST. Feb 2019 - Aug. 2024
• BSc., Computer Science, KAIST. Mar 2014 - Feb. 2019

Publications

Asterisks (*) mark equal contribution.

• Evaluating Diverse Large Language Models for Automatic and General Bug ReproductionJournal
  Sungmin Kang* and Juyeon Yoon* and Nargiz Askarbekkyzy and Shin Yoo IEEE Transactions on Software Engineering
• A Quantitative and Qualitative Evaluation of LLM-based Explainable Fault LocalizationConference
  Sungmin Kang* and Gabin An* and Shin Yoo
  The International Conference on the Foundations of Software Engineering (FSE 2024)
• The GitHub Recent Bugs Dataset for Evaluating LLM-based Debugging ApplicationsDemonstration
  Jae Yong Lee, Sungmin Kang, Juyeon Yoon, Shin Yoo
  International Conference on Software Testing, Verification and Validation (ICST), 2024
• Deceiving Humans and Machines Alike: Search-based Test Input Generation for DNNsJournal
  Sungmin Kang, Robert Feldt, Shin Yoo
  ACM Transactions on Software Engineering and Methodology (TOSEM), 2023
• SOCRATEST - Towards Autonomous Testing Agents via Conversational Large Language ModelsVision Paper
  Robert Feldt, Sungmin Kang, Juyeon Yoon, Shin Yoo
  The International Conference on Automated Software Engineering (ASE 2023), New Ideas and Emerging Results (NIER) track.
• A Bayesian Framework for Automated DebuggingConference
  Sungmin Kang* and Wonkeun Choi* and Shin Yoo
  The International Symposium on Software Testing and Analysis (ISSTA 2023)
• Towards Objective-Tailored Genetic Improvement Through Large Language ModelsVision Paper
  🏆 Best Position Paper Award, 🏆 Best Presentation Award
  Sungmin Kang and Shin Yoo
  The Genetic Improvement of Software Workshop (GI 2023)
• Large Language Models are Few-shot Testers: Exploring LLM-based General Bug ReproductionConference
  Sungmin Kang*, Juyeon Yoon*, and Shin Yoo
  The International Conference on Software Engineering (ICSE 2023)
• Arachne: Search Based Repair of Deep Neural NetworksJournal
  Jeongju Sohn*, Sungmin Kang*, and Shin Yoo
  ACM Transactions on Software Engineering and Methodology (TOSEM), 2022
• Improving Fault Localization and Automated Program Repair with Suspicious PredicatesShort Paper
  🏆 Best Short Paper Award
  Sungmin Kang and Shin Yoo
  The Korea Conference on Software Engineering (KCSE 2022)
• GLAD: Neural Predicate Synthesis to Repair Omission Faults
  Sungmin Kang and Shin Yoo
  The International Conference on Software Engineering (ICSE 2023), poster track.
• Language Models Can Prioritize Patches for Practical Program PatchingWorkshop
  Sungmin Kang and Shin Yoo
  The International Workshop on Automated Program Repair (APR 2022)
• SINVAD: Search-based Image Space Navigation for DNN Image Classifier Test Input GenerationWorkshop
  Sungmin Kang, Robert Feldt, and Shin Yoo
  The International Workshop on Search Based Software Testing (SBST 2020)
• Consistent Comic Colorization with Pixel-wise Background ClassificationWorkshop
  Sungmin Kang, Jaegul Choo, and Jaehyuk Chang
  NIPS’17 Workshop on Machine Learning for Creativity and Design

Service

Program Committee

• ISSRE | 2024
• Genetic Improvement | 2024, 2025
• DeepTest | 2025
• SBFT | 2025

Journal Reviewer

• IEEE Software | 2023
• IEEE Transactions on Software Engineering (TSE) | 2023
• ACM Transactions on Software Engineering and Methodology (TOSEM) | 2023, 2024
• Elsevier Journal of Systems and Software | 2024
• Springer Automated Software Testing | 2024
• PLOS ONE | 2024

Invited Talks

• I was invited by Georgia Tech and NUS to talk about my work on LLM-based automated debugging and future directions.
• I was invited to give a talk about explainable program repair using LLMs at the Dagstuhl Seminar 24431.
• I was invited by SureSoft and SAP to talk about LLM use in SE academia.
  • Title: “The State-of-the-art in LLM-based SE techniques: From Bug Reproduction to Program Repair”

Work Experience

• Postdoctoral Researcher, KAIST, Sep. 2024 - Feb. 2025
• Intern, Microsoft Research Asia, Sep. 2022 - Apr. 2023, under the supervision of Dr. Jian-Guang Lou.
• Intern, NAVER WEBTOON Corp, Seongnam, Jul. 2017 - Dec. 2017
• Teaching Assistant, CS206 Data Structures, Mar. 2019 - Jun. 2019
• Teaching Assistant, CS454 AI Based Software Engineering, Sep. 2019 - Dec. 2019, Sep. 2020 - Dec. 2020
  • Awarded Excellent Teaching Assistant Award for activities

Awards

• Best Presentation Award at GI 2023
• Best Position Paper Award at GI 2023
• Best Short Paper Award at KCSE 2022