Gonçalo Guiomar

Projects Artworks

Research Fellow, ETH AI Center and University of Zurich, Institute of Neuroinformatics

Previously, PhD candidate at the Champalimaud Centre for the Unknown in Lisbon under Joseph Paton

Bio

I work at the intersection of Mechanistic Interpretability and Cognition. I'm interested in how biological and artificial systems learn internal representations of the world through action and inference. My work combines computational neuroscience, reinforcement learning, and latent-state modeling to understand decision-making, reasoning, and adaptive behavior. I am currently in the labs of Valerio Mante and Benjamin Grewe. If you're interested in working with me, please check the Projects page.

Selected Publications

  1. Reasoning aligns language models to human cognition

    Gonçalo Guiomar, Elia Torre, Pehuen Moure, Victoria Shavina, Mario Giulianelli, Shih-Chii Liu, Valerio Mante (2026)

    Preprint
  2. Action suppression reveals opponent parallel control via striatal circuits

    Bruno F. Cruz, Gonçalo Guiomar, Sofia Soares, Asma Motiwala, Christian K. Machens, Joseph J. Paton (2022)

    Nature
  3. Astrophysical Constraints on the Bumblebee Model

    Gonçalo Guiomar, Jorge Paramos (2014)

    Phys. Rev. D

Selected Media

  1. Science Snapshot: Circuits of Action and Suppression

    Interview section about the work done in the Cruz, Guiomar et al Nature paper. Portuguese interview page: link.

    Interview
  2. KI im Kopf - Machen uns ChatGPT und Co. dumm? (SRF Einstein)

    Interview segment on how everyday use of generative AI can affect cognitive effort, learning behavior, and related physiological responses (German).

    Interview
  3. Metamersion: Latent Spaces (Champalimaud Foundation)

    Contributed artistic works to the exhibition, including collaborative AI-driven pieces developed with Jonathan Saldanha and the Lunar Ring collective, and supported the organization of the event as part of the Metamersion: Latent Spaces program.

    Art / Event

Teaching and Thesis Supervision