Scientists with the CMS experiment at the Large Hadron Collider searched for Higgs bosons that are produced with very large transverse momentum (pT) and then decay to a pair of W bosons. They used proton–proton collision data collected in 2016–2018, a dataset corresponding to 138 inverse femtobarns. The decay products of such a fast-moving Higgs often overlap, so the team reconstructed the visible Higgs decay as a single large-radius jet and looked at events with either one isolated lepton (electron or muon) or none.

The high-pT region is important because it tests parts of the Standard Model that are sensitive to large corrections from the strong force and to possible new effects from physics beyond the Standard Model. Earlier searches at high pT focused on other Higgs decay modes, like H→bb or H→ττ. This work is the first dedicated search for highly Lorentz-boosted Higgs bosons decaying to WW, where the two W bosons’ decay products merge into one jet when the Higgs transverse momentum is roughly 250 GeV or higher.

To pick out Higgs jets from ordinary jets produced by quarks and gluons, the analysis used a modern machine‑learning tool called PART (a transformer-based algorithm). This neural network looks at the jet’s individual particles and learns which patterns match a Higgs→WW decay. The jet tagging was further calibrated with a technique called Lund jet plane reweighting to correct subtle differences between data and simulation. The analysis split the one-lepton channel into categories aimed at different Higgs production processes: gluon fusion, vector boson fusion, and associated production with a hadronically decaying vector boson. The zero-lepton channel was treated inclusively. Signal yields were extracted by fitting the reconstructed mass of the Higgs or vector-boson candidates.

The measured value of the Higgs production rate times the H→WW branching fraction, expressed relative to the Standard Model expectation, is μ = −0.19 +0.48 −0.46. In plain terms, the result shows no evidence for a Higgs signal above the expected background. The uncertainty band is large enough that the measurement is consistent with the Standard Model prediction. The paper also reports results in the simplified template cross section framework, which helps compare different measurements of Higgs production.