Speaker
Description
The radiation pattern within high energy quark and gluon jets (jet substructure) is used extensively as a precision probe of the strong force as well as an environment for optimizing event generators for nearly all tasks in high energy particle and nuclear physics. While there has been major advances in studying jet substructure at hadron colliders, the precision achievable by collisions involving electrons is superior, as most of the complications from hadron colliders are absent. Therefore jets are analyzed which were produced in deep inelastic scattering events and recorded by the H1 detector. This measurement is unbinned and multi-dimensional, making use of machine learning to correct for detector effects. Results are presented after unfolding the data to particle level for events in the fiducial volume of momentum transfer $Q^2>150$ GeV$^2$, inelasiticity $0.2< y < 0.7$, jet transverse momentum $p_{T,jet}>10$ GeV, and jet pseudorapidity $-1<\eta_{jet}<2.5$. The jet substructure is analyzed in the form of generalized angularites, and is presented in bins of $Q^2$ and $y$. All of the available object information in the events is used to achieve the best precision through the use of graph neural networks. Training these networks was enabled by the new Perlmutter supercomputer at Berkeley Lab that has a large number of Graphical Processing Units (GPUs). The data are compared with a broad variety of predictions to illustrate the versatility of the results for downstream analyses.
arxiv:2303.13620, submitted to PLB
| Collaboration / Activity | H1 collaboration |
|---|
