CV

Research Intern at ML3 Lab

Yonsei University

• Advisor: Prof. Jaehyung Kim
• Large Language Models (LLMs)

Research Intern at MLAI Lab

KAIST AI Graduate School

• Advisor: Prof. Sungju Hwang
• Large Language Models (LLMs), Retrieval-Augmented Generation (RAG)

Research Intern at Machine Intelligence Lab

Hankuk University of Foreign Studies

• Advisor: Prof. Byunghwan Jeon
• Classification of Manufacturing products using Multi-Resnet Model

M.S.

Yonsei University

Artificial Intelligence

• Advised by Prof. Jaehyung Kim at ML3 Lab

Visiting Student

University of California, Berkeley

Computer Science

B.S.

Hankuk University of Foreign Studies

Computer Engineering

• Graduated with Honors (College of Engineering)
• Total GPA: 4.39/4.5

National Excellence Scholarship, Korea Student Aid Foundation

Academic Excellence Scholarship, Hankuk University of Foreign Studies, Korea

Align to Misalign: Automatic LLM Jailbreak with Meta-Optimized LLM Judges

International Conference on Learning Representations (ICLR)

We introduce AMIS (Align to MISalign), a bi-level optimization framework for jointly evolving jailbreak prompts and scoring templates.

Extracting and Emulsifying Cultural Explanation to Improve Multilingual Capability of LLMs

Preprint

We propose EMCEE (Extracting synthetic Multilingual Context and merging), a multilingual prompting framework that enhances LLMs on non-English queries by extracting and merging query-relevant knowledge with reasoning.

Optimizing Query Generation for Enhanced Document Retrieval in RAG

Preprint

We introduce QOQA (Query Optimization using Query expAnsion), an LLM-based query optimization method that refines RAG queries improving document retrieval quality and reducing hallucinations.