What if Artificial Intelligence could guide us in the search for new physics? That is precisely the question that has motivated a new study by IFIC that has just been published in the prestigious journal Physical Review Letters.

Artificial Intelligence is no stranger to high-energy physics; quite the opposite— it has been used for more than thirty years in the analysis of data produced at large colliders. Machines trained using machine-learning techniques, capable of analyzing and finding patterns in enormous amounts of data, are a fundamental tool in the high-energy physics toolkit. Now, a group of IFIC researchers (IFIC is a joint research center of the Spanish National Research Council and the University of València) proposes to go one step further and use artificial intelligence to explore the most promising physical models.

In the study, the team—made up of researchers from the CSIC and the University of València, Martin Hirsch, Luca Mantani, and Verónica Sanz—proposes a new strategy to analyze LHC data using genetic algorithms, an Artificial Intelligence technique inspired by natural selection. In practice, the algorithm starts with a broad set of candidate models, each representing a possible extension of the Standard Model. From there, it applies small variations (“mutations” and “combinations” among them) to generate new alternative models. In other words, the algorithm creates new theoretical models from existing ones, and each resulting model is evaluated by comparing its predictions with the available experimental data. The models that best fit the data, or that show a greater ability to improve future searches, are considered more promising. These models “survive” and are used as the basis for the next generation, while the less compatible ones are discarded. In this way, the algorithm automatically and efficiently identifies which theoretical directions have the greatest potential to reveal new physics.

This approach is especially powerful because it allows for the efficient exploration of a gigantic, multidimensional theoretical space that is impossible to scan systematically using traditional methods. Instead of analyzing each model separately—an unmanageable task due to the huge number of possible combinations—the genetic algorithm moves through the options by “jumping” among the most promising ones. By selecting, combining, and modifying only the models that show the best results, the search naturally focuses on regions of parameter space where signals of new physics are most likely to be found.

The goal is not only to better explain current data, but something more ambitious: to identify which scenarios have the greatest “discovery potential” in the near future. That is, it aims to help decide which existing models are most likely to yield truly novel contributions.

This study is particularly timely at a moment when particle physics is expected to undergo several qualitative leaps in the medium and long term, thanks to upgrades to the world’s most important colliders. In particular, the Large Hadron Collider (LHC) will see a significant increase in the intensity of its particle beams in the coming years, entering what is known as the high-luminosity phase. On the other hand, around the 2040s, the LHC will cease operation, and its successor will most likely be an even larger collider capable of venturing into unexplored territory. In this context, AI could be used as a compass to guide the search for new physics in the next generation of colliders.

The results show that this approach is capable of identifying signals that traditional analyses might overlook, opening up a new avenue for combining particle physics, advanced statistics, and artificial intelligence. All of this would help address one of the greatest problems of modern physics: the fact that the Standard Model of particle physics is incomplete. To complete it—or perhaps replace it—it is necessary to find new phenomena that allow us to discriminate among the various theoretical proposals. Artificial Intelligence tools can help us precisely by exploring this unknown territory far more efficiently and by guiding the search for new fundamental theories.