Andrea has been a PhD candidate at the Complexity Science Hub since September 2026.

She is particularly interested in belief dynamics and belief networks, and how statistical physics models can help explain real-world phenomena like radicalization and polarization; in machine behavior and the use of large language models as digital twins of human decision-making; and in identifying real-life applications of this research.

At CSH, she works within the ERC-funded Collective Minds research group. Andrea graduated in 2025 with a master’s degree in Physics of Complex Systems and Biophysics from the Universitat de Barcelona, having earned her bachelor’s degree in Physics at the same university.

Her earlier work focused on how closely large language models mirror human decision-making, replicating large-scale game-theoretic experiments to test whether open models like Llama, Mistral, and Qwen can serve as behavioral stand-ins for humans in the social sciences.