PeopleofSuperintelligence

Researcher

Cristóbal Rojas

Also known as Luis Cristóbal Rojas

Cristóbal Rojas is a mathematician and AI researcher who studies the mathematical limits and expressive power of neural networks. He earned doctorates in computer science and applied mathematics in 2008, did postdoctoral work in Toronto through 2011, and later served on the faculty of Universidad Andrés Bello. His research spans computational dynamics, computable PAC learning, graph neural networks and Transformer attention; a coauthored attention study was selected for ICLR 2026. As of October 2026, he is an associate professor at UC Chile's Institute for Mathematical and Computational Engineering and a principal investigator in CENIA's neuro-symbolic AI line.

Timeline

  1. 2008

    Doctorates in computing and mathematics

    In 2008, he earns a PhD in computer science at École Polytechnique and a PhD in applied mathematics at the University of Pisa.

  2. 2010–2011

    Postdoctoral work and faculty at UNAB

    He works as a postdoctoral researcher in Toronto in 2010–2011 and later serves on the faculty of Universidad Andrés Bello in Santiago.

  3. 2017

    Computationally hard attractors

    In 2017, he and Michael Yampolsky show that some real quadratic maps have computationally intractable attractors.

  4. 2023

    Computable PAC learning

    He coauthors a 2023 study that resolves three open questions in computable PAC learning, including one about a decidable class that requires uncomputably fast-growing sample complexity for proper learning.

  5. 2024

    One-dimensional MPNN features

    He coauthors a 2024 study that gives a simple proof that one-dimensional features can match the Weisfeiler–Leman test in message-passing graph neural networks with non-polynomial analytic activations, and validates the result experimentally.

  6. 2025

    Mathematics for neuro-symbolic AI

    In 2025, he and Pablo Barceló are principal investigators of CENIA's RL2 neuro-symbolic AI line; their work uses mathematics to study structural limits of large language models.

  7. 2026

    Transformer attention at ICLR

    In 2026, his coauthored study of positional and symbolic attention in Transformers is selected for ICLR; it defines two mutually exclusive attention-head behaviors and measures them.

  8. As of October 2026

    Associate professor and CENIA investigator

    He is an associate professor at the Institute for Mathematical and Computational Engineering at UC Chile and a principal investigator in CENIA's RL2 neuro-symbolic AI line.

Career

  1. Since 2025CENIAPrincipal Investigator
  2. NowPontifical Catholic University of ChileAssociate Professor
  3. Universidad Andrés BelloFaculty Member

Education

  1. 2008École PolytechniquePhD
  2. 2008University of PisaPhD

From the biography.

  1. The United States employs an estimated 48% of people shaping AIIncludes Pontifical Catholic University of Chile. We estimate where the people shaping AI work today, how universities, companies and research institutes split across countries, and how country shares have changed since the 1960s.
  2. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionIncludes Pontifical Catholic University of Chile. We estimate where the people shaping AI studied and where their careers led. Massachusetts Institute of Technology leads with 467 former students, including 291 recorded doctorates.
  3. The median h-index is 48 at universities and 14 at frontier AI labsIncludes École Polytechnique. The median h-index is 48 at universities and 14 at frontier AI labs. We compare publishing careers, recent citations, dual appointments and job moves across 2,420 university researchers and 308 frontier-lab researchers.
  4. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other universityIncludes École Polytechnique. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other university. We compare doctorate and 5+ year employment links, career timing and current workplaces for 187 major-prize and 1,111 best-paper winners.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Cristóbal Rojas yet. Send us a free-licensed photo.

  1. Perfiles - CENIA cenia.cl. Checked 21 October 2026.
  2. Members | Mark Braverman's Home Page mbraverm.princeton.edu. Checked 21 October 2026.
  3. Computational intractability of attractors in the real quadratic family arxiv.org. Checked 21 October 2026.
  4. Find a witness or shatter: the landscape of computable PAC learning. proceedings.mlr.press. Checked 21 October 2026.
  5. On dimensionality of feature vectors in MPNNs arxiv.org. Checked 21 October 2026.
  6. Workshop CENIA: Inteligencia Artificial al servicio de las personas cenia.cl. Checked 21 October 2026.
  7. Decoupling Positional and Symbolic Attention Behavior in Transformers arxiv.org. Checked 21 October 2026.
  8. Felipe Urrutia y Pablo Barceló participaron en el primer workshop presencial en la Universidad de Yale - CENIA cenia.cl. Checked 21 October 2026.
  9. Luis Cristóbal Rojas - Pontificia Universidad Católica de Chile uconline.uc.cl. Checked 21 October 2026.
  10. Investigación - CENIA cenia.cl. Checked 21 October 2026.