FORCE: Foundation Models for Reconfigurable Computational Efficiency in Particle Physics

The interdisciplinary research project FORCE (Foundation Models for Reconfigurable Computational Efficiency in Particle Physics) aims to evaluate Large Language Models (LLMs) with regard to their potential for applications in particle physics. The focus is on the development and systematic analysis of “foundation models” for physical processes. This includes both classification methods and the generation of physical events. A central challenge of the project is the development of suitable embeddings and tokenization strategies for continuous physical data. By analyzing and comparing different models, both small and large LLMs, their suitability for transfer and classification tasks in particle physics will be investigated.