Vinicius Dugue

Hi, I'm Vinicius — a CSULB grad building things at the intersection of graphics, simulation, and learning systems.

My background spans real-time graphics, physically based rendering, reinforcement learning, and interactive simulations in Unity. I'm drawn to projects where the underlying math becomes visible — where you can see a gradient, a light path, or a vector relationship.

Right now I'm focused on generative agents and the visual tools that make their behavior legible. The 3D embedding viewer above is a small example: instead of staring at a 300-dimensional vector, you can rotate the cluster and watch the analogy lines line up.