Celia Rubio Madrigal
CISPA Helmholtz Center for Information Security, Germany
Celia Rubio Madrigal is a mathematician and computer scientist from Spain who works on graph machine learning. She studies how structures shape data representation and computation, with broader interests in generalization trade-offs and architectural efficiency.
the artwork
Cambios latentes, conexiones simples
Spanish - Latent changes, simple connections - 2026 - digital collage, photography
audio commentary
This is my dog on a videocall with me. I was inspired by Picasso’s Le Taureau, where a bull is simplified until only its essence remains. I work on simplifying graph ML architectures, so I share that goal too. I represented my dog’s aging and our relationship now that I live abroad for my PhD. Around him are graph-like photographs I took on conference trips over the past two years. Whenever I return home, my dog seems to have suddenly aged, but the essence of our bond remains. In the end, simple connections are what matter, despite the latent changes that occur when I am not there to see them.
Celia Rubio Madrigal
Gradual changes in loved ones perceived as drastic after long periods apart
mathematics in Spanish culture
I was also inspired by a depiction of hounds on a 15th-century olambrilla mudéjar (Hispano-Moresque floor tile) from Toledo in Spain, the region where my family is from. The tile-making tradition of medieval Al-Andalus lasted for centuries in the Iberian Peninsula. Many tileworks, such as those in the Alhambra palace, show geometric patterns with complex symmetries. These can be described through mathematical tessellations. The way they are constructed is an example of algorithmic art, an art form I greatly enjoy and have also used here.
Celia Rubio Madrigal
