Speaker
Dr
Sofia Vallecorsa
(CERN)
Description
This project addresses the growing need for scalable and efficient detector simulation in HEP, leveraging self-supervised learning and generative modeling to enable fast, generalizable simulations. By learning from unlabeled data and exploiting the intrinsic structure of detector responses, the proposed approach aims to reduce simulation time while maintaining physical fidelity. The project focuses on developing a unified model capable of adapting to various detector configurations, thereby eliminating the need for retraining for each setup. Expected outcomes include enhanced generalization across experimental conditions and flexible reusable architectures.
CERN group/ Experiment
IT-FTI
| Working area | Area 1" Cutting Edge AI for Offline Data Processing |
|---|---|
| Project goals | - To design a self-supervised learning framework tailored for detector simulation tasks. - To train generative models capable of reproducing realistic detector responses. - To explore the generalization capabilities of the trained models across different detector geometries and experimental conditions |
| Timeline | Y1: Foundation and representation learning (train the initial architecture for an initial set of of use case) Y2: Generalization and performance benchmarking (extend the range of use cases and evaluate the limitations and requirements in term of data / computational resources) |
| Available person power | Staff (supervision) 1 origin (externally funded) |
| Additional person power request | 1 quest 1 doct |
| Is this an already ongoing activity? | Yes |
| Indicative hardware resources needs | Access to a 4xA100 node. Access to HPC or cloud resources for large scale tests. |
Author
Dr
Sofia Vallecorsa
(CERN)