Skip to content

Edinburgh Summer School on Generative AI for Extremes

The University of Edinburgh GAMEX
8–11 September 2026
Edinburgh, UK

Overview

Building on the success of GAME 2025 in Edinburgh, the first worldwide workshop on generative AI modelling for extreme events, this summer school is led by GLE2N (Glasgow–Edinburgh Extremes Network) and brings PhD students and early-career researchers to the interface of generative AI and extreme value theory. It is an official ECAS course (European Courses in Advanced Statistics), and there is no registration fee.

Due to exceptionally high interest, registration for the Edinburgh Summer School on Generative AI for Extremes has now closed.

To be considered if additional places become available, or to join the waiting list, please complete the Expression of Interest form. We will contact you if a place becomes available.

Expression of Interest

Programme

Date Theme Agenda
Tuesday
8 Sep
Foundations and Preparations for Generative Modelling Registration and welcome coffee; lectures; practical session
Wednesday
9 Sep
Generative Modelling for Extremes and Risk Lectures; practical session; group photo; open practice and office hours
Thursday
10 Sep
Multivariate Extremes, Cascades, and Simulation Lectures; practical session; open practice and office hours
Friday
11 Sep
Other Frontier Topics Lectures; closing remarks
Summer School participants and instructors outside the James Clerk Maxwell Building
Group Photo.

Chair

Professor Miguel de Carvalho
Chair of Statistics & Data Science
Elected Fellow of Generative AI Laboratory

Organisers

Johnny Lee (Edinburgh), Lambert de Monte (Edinburgh), Jordan Richards (Edinburgh), Daniela Castro-Camilo (Glasgow), Mengran Li (Glasgow), Teng Wei Yeo (Glasgow), and Luca Trapin (Bologna).

Materials

Materials will be made available by the instructors in due time. Other resources include selected parts of the Handbook of Statistics of Extremes and parts of the recent special issue on Bridging Heavy Tails and AI.

Participants should have Python and Jupyter available; examples of required modules include torch, numpy, scipy, pandas, and matplotlib.

Venue

The summer school will take place at JCMB (James Clerk Maxwell Building), University of Edinburgh, Peter Guthrie Tait Road, King’s Buildings, Edinburgh EH9 3FD.

Rooms: Lecture Theatre B for lectures and Teaching Studio 3217 for practical sessions.

James Clerk Maxwell Building
JCMB, University of Edinburgh.

Accommodation

Participants are responsible for arranging their own accommodation. Options around the King’s Buildings campus include Glendale Guest House and other guest houses around Mayfield Road, Minto Street, and Newington. Potentially affordable options near King’s Buildings include Allison House 16, Allison House 17, Ardgowan Guest House, JOIVY Meadows Park En-suite Rooms, Travelodge Edinburgh Cameron Toll, and Tantallon Bed and Breakfast; near the city centre, options include Travelodge Edinburgh Central Waterloo Place, Hub by Premier Inn Edinburgh Royal Mile Hotel, and Premier Inn Edinburgh Central (Lauriston Place) Hotel. The University of Edinburgh’s Summer Stays booking page lists additional accommodation options during the summer period.

Getting to JCMB

From Edinburgh Waverley train station, JCMB is around 3 miles south at the King’s Buildings campus. Allow around 20–30 minutes by bus or taxi, depending on traffic.

Map from Edinburgh Waverley to JCMB

Open route in Google Maps

Partners