Data-driven discovery strategy for standard model effective field theory searches

Abstract

We present a novel strategy to uncover indirect signs of new physics in collider data using the Standard Model Effective Field Theory (SMEFT) framework, offering notably improved sensitivity compared to traditional global analyses. Our approach leverages genetic algorithms to efficiently navigate the high-dimensional space of operator subsets, identifying deformations that improve agreement with data without relying on prior UV assumptions. This enables the systematic detection of SMEFT scenarios that outperform the Standard Model in explaining observed deviations. We validate the approach on current LHC and LEP measurements, perform closure tests with injected UV signals, and assess performance under high-luminosity projections. The algorithm successfully recovers relevant operator subsets and highlights directions in parameter space where deviations are most likely to emerge. Our results demonstrate the potential of SMEFT-based discovery searches driven by model selection, providing a scalable framework for future data analyses.

Figures (3)

• Figure 1: Bar plot showing the pseudo-posterior probabilities of the 8 top-ranked models based on their BIC scores, in a closure test with L1 SM-generated pseudodata. The SM emerges as the most probable model, with its dominance increasing at the HL-LHC as expected from improved precision.
• Figure 2: Left: Same as Fig. \ref{['fig:sm-closure']}, but for BSM-injected pseudodata. At HL-LHC, two injected operators are reliably recovered, while the Yukawa-like one remains elusive due to degeneracies. At LHC precision, the SM is disfavoured, yet the characterization of NP remains inconclusive. Right: Comparison of the 95% confidence intervals for selected operators between the top three scoring models and the global fit, using the HL-LHC dataset. The result highlights that the global fit fails to accurately characterize the BSM signal.
• Figure 3: Same as Fig. \ref{['fig:sm-closure']}, using real LEP + LHC data. Results are shown both with and without RGE, highlighting its impact on model selection. Including RGE reduces the statistical preference for the top models, since operator mixing allows deviations to be accommodated by a broader set of SMEFT contributions. The SM does not appear in the top 8 models and is therefore not shown, but it remains statistically competitive and is not excluded.