Minsuk Sung · 성민석
Curriculum Vitae연구 이력 (CV)
Ph.D. Student in Artificial Intelligence인공지능학과 박사과정 · Korea University고려대학교
Research interests
Human & AI Visual Perception · Explainable AI · Vision-Language Models · AI in Finance
Education
- Ph.D. student in Artificial Intelligence, Korea University — Mar. 2020–Present
- B.S. in Computer Engineering, Hongik University — Mar. 2014–Feb. 2020
Research experience
- Multi-Agent Quant Lab (MAQ Lab) — Lab leader, Nov. 2025–Present
Open study building live-ready algorithmic trading systems with AI coding agents; two cohorts completed. Papers at ICAIF 2026 and MARBLE 2026. Website - Cognitive Systems Lab — Nov. 2020–Present
Human-like object detection. Advisor: Christian Wallraven. - PRML Lab — Jan. 2020–Oct. 2020
Reinforcement learning for optimal trade execution and slippage minimization.
Work experience
- Eugene Investment & Securities — Intern, CM Team, Bond Solutions Office, Fixed Income Division — May 2026–Present
Data pipelines and AI tooling for the fixed-income business. - Mirae Asset Daewoo — Intern — Jan.–Feb. 2021
Financial app reputation analysis and comparative NLP research.
Publications
2026
- Minsuk Sung, Doohwi Cha, Juyeong Lee, Minjae Lee, Donghwa Seo, Seunghan Son, Yoontae Hwang. “AlphaLeak: What ‘Blind’ LLM Trading Benchmarks Still See.” ICAIF 2026 (ACM International Conference on AI in Finance), accepted (oral).
- Doohwi Cha, Minsuk Sung, Seunghan Son, Juyeong Lee, Donghwa Seo, Minjae Lee, Hyeonjun Yeo, Yoontae Hwang. “CallRank: Isolating What Changed in Earnings-Call Q&A for Sector Ranking and Cost-Adjusted Alpha.” ICAIF 2026, accepted (oral).
- Juyeong Lee, Donghwa Seo, Minjae Lee, Seunghan Son, Minsuk Sung, Doohwi Cha, Yoontae Hwang. “Decision-Focused Learning of the Gerber Threshold.” ICAIF 2026, accepted (oral).
- Seunghan Son, Doohwi Cha, Minjae Lee, Juyeong Lee, Minsuk Sung, Donghwa Seo, Yoontae Hwang. “SlipClock: A Conditional Displayed-Depth Cost Benchmark for Evaluating Financial AI Strategies in Crypto Futures.” ICAIF 2026, accepted.
- Doohwi Cha, Hyeonjun Yeo, Gyuil Jung, Jeongkyoo You, Minsuk Sung, Donghwa Seo, Minjae Lee, Seunghan Son, Juyeong Lee, Yoontae Hwang. “Metropolitan Housing Signals for Treasury Duration Risk Management.” ICAIF 2026, accepted.
- “The Division of Research: Information Access and Independent-Book Capacity in LLM Investment Teams.” ICAIF 2026, accepted.
- Doohwi Cha, Minsuk Sung, Minjae Lee, Hyeonjun Yeo, Hanhee Kim, Seunghan Son, Donghwa Seo, Juyeong Lee. “NEST: Stale-Robust Tokenized-Stock Measurement for a Crypto-Treasury LETF Stack.” MARBLE 2026 (International Conference on Mathematical Research for Blockchain Economy), Springer Lecture Notes in Operations Research.
- Doohwi Cha, Hanhee Kim, Hyeonjun Yeo, Minsuk Sung, Minjae Lee, Seunghan Son, Donghwa Seo, Juyeong Lee. “Pricing the Unprovable: Ambiguity, Turnstile Runs, and the Verifiability Discount in Crypto Protocols.” MARBLE 2026.
- Donghwa Seo, Doohwi Cha, Seunghan Son, Juyeong Lee, Minsuk Sung, Minjae Lee. “Basis-Only Funding Is Not Dividend Fair: Settlement Design in Crypto Equity Perpetuals.” MARBLE 2026.
- Donghwa Seo, Doohwi Cha, Seunghan Son, Juyeong Lee, Minjae Lee, Minsuk Sung. “When Cross-Venue Agreement Is Not Price Discovery: Disclosure Frontiers for 24/7 Equity-Perpetual Oracles.” MARBLE 2026.
- Youhyun Sung, Minsuk Sung, Claus-Christian Carbon, Christian Wallraven. “Art Wins: Higher Valuation Responses to Op Art Compared to Matched Optical Illusions.” ECVP 2026 (48th European Conference on Visual Perception), poster.
- Youhyun Sung, Minsuk Sung, Claus-Christian Carbon, Christian Wallraven. “Does Spatial Ambiguity Contribute to Aesthetic Preference? Evidence from Op Art and Matched Control Stimuli.” VSAC 2026 (Visual Science of Art Conference), oral presentation; abstract to appear in Art & Perception.
2019
- Ha Yoon Song, Moo Sang Baek, Minsuk Sung. “Generating Human Mobility Route Based on Generative Adversarial Network.” FedCSIS 2019, Annals of Computer Science and Information Systems, vol. 18, pp. 91–99. DOI
- Minsuk Sung, Ha Yoon Song. “Prediction on Visited Stations through the Analysis of Individual Movements Data.” Korean Institute of Information Scientists and Engineers (KIISE), 2019.
Teaching & service
- NAVER Boostcamp, cohorts 1–8 — 2021–Present: Teaching assistant for competition development and lecture content review
- Intel Korea — 2019–2023: Object detection, image classification, and OpenVINO conversion workshops
- SK company-wide AI training for new employees — Jan. 2021: Teaching assistant for machine learning and deep learning fundamentals
- National Information Society Agency (NIA) — Analytics infrastructure training
- FAST CAMPUS — Teaching assistant for course development
- Samsung Electronics DS — Deep learning leadership training
Activities
- BOAZ student data science community, organizing team — Jan. 2019–Jan. 2020: Introductory crawling and deep learning lectures, and Kaggle workshops
- HIARC, Hongik University algorithm club — Mar. 2018–Feb. 2019
- Data science training and study groups, including Hands-On Machine Learning and Deep Learning from Scratch
Awards
- Feb. 2021 — Oral disease medical imaging AI competition: 3rd place
- Dec. 2020 — 4th Big Data Festival: Excellence Award
- June 2020 — Biological optical data AI competition: 7th of 696
- Apr. 2020 — Public data temperature estimation AI competition: 2nd of 1,081
- Nov. 2019 — 3rd Big Data Festival: Gold Award
- Nov. 2019 — 7th Big Contest: Top Excellence Award
- Sept. 2019 — Space weather disaster prediction AI competition: 7th in the finals
- Mar. 2019 — Seoul Hardware Hackathon: 3rd place
Patents
- 2018 — Route optimization using delivery hubs
- 2019 — Human mobility route prediction using GANs
Contact
minsuksung@korea.ac.kr · Seoul, South Korea
연구 관심사
인간과 AI의 시각인지 · 설명 가능한 AI · 비전-언어 모델 · 금융 AI
학력
- 고려대학교 인공지능학과 박사과정 — 2020.03 ~ 현재
- 홍익대학교 컴퓨터공학과 학사 — 2014.03 ~ 2020.02
연구 경력
- Multi-Agent Quant Lab (MAQ Lab) — 랩장, 2025.11 ~ 현재
AI 코딩 에이전트로 실거래 가능한 알고리즘 트레이딩 시스템을 만드는 오픈 스터디, 2기 운영 완료. ICAIF 2026·MARBLE 2026 논문 발표. 홈페이지 - Cognitive Systems Lab — 2020.11 ~ 현재
인간 유사 객체 탐지 연구. 지도교수: Christian Wallraven. - PRML Lab — 2020.01 ~ 2020.10
강화학습 기반 최적 매매 집행 및 슬리피지 최소화.
경력
- 유진투자증권 — 채권금융본부 채권솔루션실 CM팀 인턴 — 2026.05 ~ 현재
채권 비즈니스를 위한 데이터 파이프라인과 AI 도구 개발. - 미래에셋대우 — 인턴 — 2021.01 ~ 2021.02
금융 앱 평판 분석 및 NLP 비교 연구.
논문
2026
- Minsuk Sung, Doohwi Cha, Juyeong Lee, Minjae Lee, Donghwa Seo, Seunghan Son, Yoontae Hwang. “AlphaLeak: What ‘Blind’ LLM Trading Benchmarks Still See.” ICAIF 2026 (ACM International Conference on AI in Finance), 채택 (구두 발표).
- Doohwi Cha, Minsuk Sung, Seunghan Son, Juyeong Lee, Donghwa Seo, Minjae Lee, Hyeonjun Yeo, Yoontae Hwang. “CallRank: Isolating What Changed in Earnings-Call Q&A for Sector Ranking and Cost-Adjusted Alpha.” ICAIF 2026, 채택 (구두 발표).
- Juyeong Lee, Donghwa Seo, Minjae Lee, Seunghan Son, Minsuk Sung, Doohwi Cha, Yoontae Hwang. “Decision-Focused Learning of the Gerber Threshold.” ICAIF 2026, 채택 (구두 발표).
- Seunghan Son, Doohwi Cha, Minjae Lee, Juyeong Lee, Minsuk Sung, Donghwa Seo, Yoontae Hwang. “SlipClock: A Conditional Displayed-Depth Cost Benchmark for Evaluating Financial AI Strategies in Crypto Futures.” ICAIF 2026, 채택.
- Doohwi Cha, Hyeonjun Yeo, Gyuil Jung, Jeongkyoo You, Minsuk Sung, Donghwa Seo, Minjae Lee, Seunghan Son, Juyeong Lee, Yoontae Hwang. “Metropolitan Housing Signals for Treasury Duration Risk Management.” ICAIF 2026, 채택.
- “The Division of Research: Information Access and Independent-Book Capacity in LLM Investment Teams.” ICAIF 2026, 채택.
- Doohwi Cha, Minsuk Sung, Minjae Lee, Hyeonjun Yeo, Hanhee Kim, Seunghan Son, Donghwa Seo, Juyeong Lee. “NEST: Stale-Robust Tokenized-Stock Measurement for a Crypto-Treasury LETF Stack.” MARBLE 2026 (International Conference on Mathematical Research for Blockchain Economy), Springer Lecture Notes in Operations Research.
- Doohwi Cha, Hanhee Kim, Hyeonjun Yeo, Minsuk Sung, Minjae Lee, Seunghan Son, Donghwa Seo, Juyeong Lee. “Pricing the Unprovable: Ambiguity, Turnstile Runs, and the Verifiability Discount in Crypto Protocols.” MARBLE 2026.
- Donghwa Seo, Doohwi Cha, Seunghan Son, Juyeong Lee, Minsuk Sung, Minjae Lee. “Basis-Only Funding Is Not Dividend Fair: Settlement Design in Crypto Equity Perpetuals.” MARBLE 2026.
- Donghwa Seo, Doohwi Cha, Seunghan Son, Juyeong Lee, Minjae Lee, Minsuk Sung. “When Cross-Venue Agreement Is Not Price Discovery: Disclosure Frontiers for 24/7 Equity-Perpetual Oracles.” MARBLE 2026.
- Youhyun Sung, Minsuk Sung, Claus-Christian Carbon, Christian Wallraven. “Art Wins: Higher Valuation Responses to Op Art Compared to Matched Optical Illusions.” ECVP 2026 (48th European Conference on Visual Perception), 포스터.
- Youhyun Sung, Minsuk Sung, Claus-Christian Carbon, Christian Wallraven. “Does Spatial Ambiguity Contribute to Aesthetic Preference? Evidence from Op Art and Matched Control Stimuli.” VSAC 2026 (Visual Science of Art Conference), 구두 발표; 초록은 Art & Perception에 게재 예정.
2019
- Ha Yoon Song, Moo Sang Baek, Minsuk Sung. “Generating Human Mobility Route Based on Generative Adversarial Network.” FedCSIS 2019, Annals of Computer Science and Information Systems, vol. 18, pp. 91–99. DOI
- Minsuk Sung, Ha Yoon Song. “Prediction on Visited Stations through the Analysis of Individual Movements Data.” 한국정보과학회 2019.
강의 및 교육
- NAVER 부스트캠프 1기~8기 — 2021 ~ 현재: 대회 제작 및 강의 내용 검수 조교
- Intel Korea — 2019 ~ 2023: 객체 탐지 및 이미지 분류 교육, OpenVINO 변환 실습
- SK 인공지능 신입 전사 교육 — 2021.01: 머신러닝·딥러닝 기초 교육 조교
- 한국정보화진흥원(NIA) — 분석인프라 교육
- FAST CAMPUS — 강의 제작 조교
- 삼성전자 DS — 딥러닝 리더 교육
활동
- BOAZ 연합 동아리 운영진 — 2019.01 ~ 2020.01: 크롤링·딥러닝 입문 강의와 Kaggle 실습
- 홍익대 알고리즘 동아리 HIARC — 2018.03 ~ 2019.02
- 데이터 캠퍼스 교육 및 스터디: Hands-On ML, Deep Learning from Scratch 등
수상
- 2021.02 — 구강계질환 의료영상 AI 경진대회 3등
- 2020.12 — 제4회 빅데이터 페스티벌 우수상
- 2020.06 — 생체 광학 데이터 분석 AI 분석대회 7등 / 696
- 2020.04 — 공공데이터 활용 온도 추정 AI 경진대회 2등 / 1,081
- 2019.11 — 제3회 빅데이터 페스티벌 금상
- 2019.11 — 제7회 빅콘테스트 최우수상
- 2019.09 — 우주전파재난 예측 AI 경진대회 본선 7등
- 2019.03 — 서울 하드웨어 해커톤 3등
특허
- 2018 — 배송 허브를 활용한 경로 최적화
- 2019 — GAN을 이용한 인간 이동 경로 예측
연락처
minsuksung@korea.ac.kr · 대한민국 서울