Trained on billions of simulated proton-proton collisions from the Large Hadron Collider (LHC), the system aims to streamline data reconstruction, identify particles, flag anomalies, and optimize future detector designs. This technological leap addresses the growing complexities of data analysis as CERN prepares for the High-Luminosity LHC, which will increase collision volumes tenfold by 2030.
“High-energy physics provides a unique environment in which artificial intelligence can be developed and tested with exceptional rigour. We are partnering with Mistral, a leading European AI company, to advance AI capabilities that can serve the whole physics community and be shared,” said Enrica Porcari, CERN's Chief Information Officer.
The collaboration represents a structural shift in how researchers process experimental output. Today, event reconstruction relies on a sequential chain of specialized algorithms that progressively translate raw detector signals into physical objects suitable for analysis. While machine learning has long assisted with isolated tasks within this pipeline, the proposed foundation models offer the potential to learn across the entire reconstruction chain simultaneously, supporting multiple workflows within a single architecture.
CERN currently houses approximately one exabyte of experimental data generated by the LHC. With the upcoming transition to the High-Luminosity LHC around 2030, the rate of particle collisions is projected to surge by an order of magnitude. This unprecedented volume and precision create a rigorous proving ground for advanced machine learning systems.
“We are proud to work with CERN to explore what large-scale AI models can bring to the growing complexity of particle physics. Together, we have the opportunity to push the frontiers of high-energy physics, developing AI capabilities that have the potential to lead to new scientific discoveries,” said Johannes Brandstetter, Vice-President of AI for Engineering at Mistral.