Skip to content

Project

The project brings together complementary expertise in Generative Artificial Intelligence (AI) and Extreme Value Theory (EVT) through coordinated research, training, and network-building activities, implemented over the period January–December 2026.

Partners

logos

Motivation

Extreme events such as floods, heatwaves, and financial crises are increasingly central to scientific analysis and decision-making. While modern generative models scale efficiently to complex, high-dimensional data, they often fail to reproduce tail behaviour and extreme dependence structures. GAMEX connects the scalability of Generative AI with the rigor and interpretability of EVT by fostering collaboration between experts in EVT and Generative AI, with the goal of developing models that are both scalable and theoretically grounded.

Objectives

  • Scientific advancement: strengthen collaboration on methods that integrate Generative AI and EVT for modeling extremes.
  • Community building and training: support early-career researchers through seminars, events, and short-term visits.
  • Strategic development: expand the network with additional academic and industrial partners and lay foundations for future funding initiatives (e.g., MSCA Staff Exchange, Doctoral Networks).