Evaluating the Impact of Detector Design on Jet Flavor Tagging for Future Colliders

Abstract

Jet flavor tagging is of utmost importance for unlocking the full physics potential of any future collider experiment. The performance of any jet flavor identification algorithm depends both on its underlying architecture and on the detector's design and capabilities. In this work, we present an analysis of the dependence of jet tagging algorithm performance on three detector designs being considered for future $e^{+}e^{-}$ colliders. To fully exploit the potential of these detector concepts, we utilize a graph neural network-based jet tagging algorithm. In addition, we evaluate the impact on the jet tagging performance of variations in the tracking and calorimeter systems for one of these detector concepts, the SiD detector, as well as the dependence on the center-of-mass energy.