Searching for Extreme Events in Multi-lepton Data from the LHC

The Standard Model particles cannot represent the complete set of nature's constituents, but there's no guarantee that new particles to be discovered would be light enough to be produced on-shell at the LHC. Thus, indirect methods of probing higher mass scales become increasingly interesting in the search for new physics at the energy frontier. Effective Field Theory (EFT) is an example of such an indirect probe, which offers a model-independent method of extending the discovery reach of the LHC. As part of the EFT analyses, data from the CMS detector is explored. The full Run 2 data is preselected to be top production events with multiple leptons in their final states. The multi-lepton data are then classified in dataframe according to several characteristics (e.g. jet multiplicity). Top high-energy events in each class were searched. The observed data were compared with the simulation data of the EFT model qualitatively. A processor is written under the Coffea framework to run large quantities of data at scale using distributed batch system HTCondor with several hundreds of cores concurrently. A bottleneck of the computation performance is identified as the data transfer of input files (XRootD servers).