Daniel Murnane

TL;DR: This work demonstrates the applicability of GNNs to these two diverse particle reconstruction problems, which have unique challenges and characteristics, but both have high dimensionality, high degree of sparsity, and complex geometric layouts.

TL;DR: CollaboratorBit as mentioned in this paper is a new code for the calculation of high energy collider observables in theories of physics beyond the Standard Model (BSM) ColliderBit features a generic interface to BSM models, a unique parallelised Monte Carlo event generation scheme suitable for large-scale supercomputer applications, and a number of LHC analyses, covering a reasonable range of the BSM signatures currently sought by ATLAS and CMS.

TL;DR: CollaboratorBit as mentioned in this paper is a new code for the calculation of high energy collider observables in theories of physics beyond the Standard Model (BSM), which can be used to calculate the likelihoods for Higgs sector observables and LEP searches for BSM particles.

TL;DR: A suite of extensions to the original model of graph neural networks, with encouraging results for hitgraph classification are presented, and increased performance is explored by constructing graphs from learned representations which contain non-linear metric structure, allowing for efficient clustering and neighborhood queries of data points.

TL;DR: The Exa.TrkX project as mentioned in this paper applied geometric learning concepts such as metric learning and graph neural networks to HEP particle tracking and achieved tracking efficiency and purity similar to production tracking algorithms.