Juhyun Oh

Juhyun Oh

Hi! I am a Ph.D. candidate working on Natural Language Processing at KAIST, advised by Alice Oh. Prior to this, I received my M.S. degree in Linguistics at Seoul National University.

My research focuses on understanding the capabilities and limitations of language models, and examining how they differ from human. In 'Uncovering Factor Level Preferences,' I explore the reasons behind human-model preference divergence. In 'Generative AI Paradox in Evaluation,' I demonstrate the disconnect between LLMs' generative abilities and their reliability as evaluators. Over the past few years, I have contributed to the NLP community by developing benchmarks for evaluating language models on linguistic tasks, including NLI and offensive language detection for Korean.

Email: 411juhyun [at] kaist.ac.kr

Links: [Google Scholar] [Twitter] [CV]

Selected Publications

For a complete list, check my Google Scholar

Teaching Experience

  • (Spring, 2024) Teaching Assistant @ AI Ethics (KAIST)