Status
I am a Master of Science in Intelligent Information Systems (MIIS) student at Carnegie Mellon University, focusing on designing and analyzing LLM agents that behave robustly and transparently in human-facing environments.

Current Research
I am currently advised by Prof. Graham Neubig in CMU, focusing LLMs as coding agents. My recent research includes PR Arena, a platform for evaluating and benchmarking agentic coding assistants through paired pull request (PR) generations, and SYCON Bench, a benchmark for analyzing sycophancy in multi-turn dialogues. I am also working with Prof. Jinho D. Choi in Emory University, designing an educational conversational AI, 🧚 Tinker Tales, and developing an agentic method for long-context instruction following capabilities.

Prospective Research Interest
I am interested in:

• Methodologies that elicit deep reasoning capabilities in language models by controlling latent variables (e.g., Quiet-STaR, Coconut).
• Coding agents in human-facing environments, such as agents with effective human-feedback loop or unverifiable metrics to evaluate code patches.

Academic Milestone
I received my Bachelor's degree in Computer Science from Korea Advanced Institute of Science and Technology (KAIST), where I studied machine learning, NLP, and human-computer interaction. I'm very grateful to Prof. Minsuk Kang, Prof. Jeehoon Kang, and NC Soft mentor Hochang Lee for their guidance.

View my Resume