César Miguel Valdez Córdova

I study how structure emerges in learning systems, both biological and artificial. When we apply methods like dimensionality reduction or train neural networks on data, we're making implicit choices about what patterns matter. I'm interested in understanding how the shape of the data itself, its underlying geometry, guides these discoveries. I'm developing tools and frameworks for agentic AI systems that can reason geometrically about scientific data, learning to make principled analytical choices based on the intrinsic structure of the systems they're analyzing.