I’m a Ph.D. Candidate @ Music and Audio Computing Lab, KAIST.

I am a machine learning researcher focusing on representation learning and controllable generative modeling for sequential and structured data. My previous work has mainly focused on symbolic music data, exploring various structured representations such as tokenized sequences, matrices, and graphs for music generation and understanding. I have hands-on experience in dataset construction, preprocessing, and model development across the entire ML pipeline, particularly with Transformer- and Diffusion-based architectures. Ultimately, my long-term goal is to enable machines to learn and understand the underlying structures and patterns in human-created data.