What is this event?
The AISSAI Hackathon on AI for Particle Accelerators is an intensive five-day collaborative event bringing together 30–50 researchers — accelerator physicists, AI scientists, and HPC engineers — to work hands-on on concrete scientific challenges at the frontier of machine learning and particle accelerator science.
Participants will form small competing teams, share access to dedicated HPC resources at Jean-Zay (IDRIS), and tackle three major challenge tracks:
- Surrogate Modeling — building fast, reliable emulators of complex accelerator subsystems
- Anomaly Detection & classification — detecting, classifying, and explaining abnormal machine behavior
- Control, Optimization & Decision Support — AI-assisted tuning and operational intelligence
Each track will feature two teams adopting contrasting methodological approaches, with a final plenary confrontation of results on Friday afternoon.
The event is accompanied by resident artists who will document the week through observation and produce an illustrated comic deliverable.
Who should apply? PhD students, postdoctoral researchers, and early-career engineers from the accelerator physics, physics, AI/ML, and HPC communities. Senior scientists are welcome. Participation is by selection— the organizing committee will assemble teams based on skill profiles and scientific preferences collected at registration. Attendance is free of charge. A limited number of travel and accommodation grants are available for early-career participants.