Speaker
Description
Determining the shape of the Higgs potential is essential for assessing the stability of the vacuum, establishing how electroweak symmetry is broken, and searching for new physics. The measurement of the trilinear Higgs self-coupling lambda_HHH is the key, and Higgs boson pair production (HH) is the only process with direct sensitivity to it.
The HH to 4b final state, in which both Higgs bosons decay to b-quark pairs, has the largest branching ratio and is one of the most sensitive channels.
Improving the performance of b-jet identification (b-tagging) is essential for increasing the sensitivity of this analysis. At the ATLAS experiment, the introduction of transformers and related algorithms has substantially improved b-jet identification in the offline analysis. Collecting HH to 4b events with high efficiency requires event selection with b-jet triggers at the online stage. Because of latency constraints, the online environment and the offline analysis differ in the b-tagging algorithms that can be used and in the detector information that is available. Scale factors (SFs) that correct the differences in identification efficiency between data and simulation must therefore be derived, taking proper account of the differences between the online and offline efficiencies and of their correlation.
In this work, the search sensitivity to non-resonant HH to 4b is evaluated with these trigger SFs taken into account, and the impact of the statistical and systematic uncertainties of the SFs on the sensitivity and on the allowed region of lambda_HHH is assessed. This poster shows the impact of these trigger-calibration results on the search sensitivity.