Huiwon Jang

Huiwon Jang is a PhD student in AI at KAIST and a project and research lead at RLWRLD, where she co-leads the development of RLDX-1, a foundation model for dexterous robot hands. She is the primary author of ContextVLA, a vision-language-action model framework that compresses multi-frame observations into compact representations for efficient temporal reasoning in robotic tasks. Her research spans vision-language models, diffusion transformers, and representation learning for robotics.