About

Hi, I'm Hyeonbin.

I’m a Master’s student at KAIST AI advised by Minjoon Seo.

My grad school journey started with using LLM reasoning as a path toward AGI. However, I soon ran into the "bitter lesson": models often feign competence by memorizing training data, causing performance to plateau the moment they step outside that distribution.

My work now focuses on breaking through that ceiling. I’m interested in designing learning signals that push models beyond simple pattern matching and verifying those capabilities in rigorous robotics testbeds.

With this end, I'm currently interning at ALIN Lab, where I’m researching Vision-Language-Action Models.

I’m always open to collaborations. Feel free to reach out!