I’m a young researcher at the intersection of machine learning, robotics, and control theory. My current focus is on combining model-based and learning-based control — not just to improve performance, but to build systems that understand the physical world.
I’m fascinated by phenomena like Hamiltonians, Lagrangians, Lyapunov theory, and even quantum mechanics. These are not just math tools to me; they’re clues to how intelligent systems should reason. I believe AGI will need strong dynamical priors — not just data — to truly grasp the world. To learn anything, an agent must predict—and be changed by its prediction errors.
This belief drives my interest in unconventional ML approaches like dynamic and liquid neural networks, energy-based models, and predictive coding. General models like Vision-Language-Action models (VLAs) can solve control problems. But I believe we're still scratching the surface. The best is ahead, and I want to be part of building it.
I plan to pursue a PhD after undergrad, to keep learning, building, and contributing to the path toward real-world intelligent systems.
