CV

Researcher

Optimization & Machine Learning Laboratory

• Advisor: Chulhee Yun
• Developed a four-term risk decomposition and compute-optimal exponents for one-pass signSGD in PLRF.

Researcher

Cryptography Laboratory

• Advisor: Junghee Cheon
• Proposed an algorithm to enhance the efficiency of CKKS bootstrapping.

Bachelor of Science

Seoul National University

Mathematics and Computer Science

High School Diploma

Seoul Science High School

Simon Marais Mathematics Competition (Pairs)

Pairs 7th place (7/448 = 2%).

Korean Undergraduates Mathematics Competition

Gold Prize.

National Cryptography Contest

Special Prize (related to OverModRaise).

Simon Marais Mathematics Competition (Individual)

Individual 8th place (8/442 = 2%).

Dean's List (Department of Natural Science)

Seoul National University

Korea Presidential Science Scholarship

Scholarship (9k/year) for 2023-2027.

Scaling Laws of SignSGD in Linear Regression: When Does It Outperform SGD?

International Conference on Learning Representations (ICLR)

Derived scaling-law theory for one-pass signSGD in the PLRF model, including risk decomposition and compute-optimal exponents.

OverModRaise: Reducing Modulus Consumption of CKKS Bootstrapping

IACR Communications in Cryptology

Proposed and implemented an algorithm improving the efficiency of CKKS bootstrapping by reducing modulus consumption.