Speaker
Description
Final states containing isolated electrons and photons (e/γ) have played a central
role in physics discoveries at the Large Hadron Collider (LHC) and will remain
vital to the ATLAS trigger strategy throughout the High-Luminosity LHC (HL-LHC)
programme. However, the substantially increased pile-up expected at the HL-LHC
will make it increasingly challenging to preserve high trigger acceptance for
these signatures while maintaining rates within the available budget.
To address this challenge, the ATLAS Phase-II trigger upgrade will introduce
the Level-0 Global Trigger into the hardware trigger chain, enabling more
sophisticated selection and refinement algorithms, albeit with a latency
constraint of a few microseconds. We present a fast, quantised convolutional
neural network (CNN) for isolated e/γ selection in the ATLAS Level-0 Global
Trigger. The network refines primitive e/γ candidates identified by the e/γ
Feature Extractor by exploiting transverse-energy and shower-shape information
from calorimeter towers in the neighbourhood of each candidate.
The performance of the algorithm is evaluated using simulated collision events
under HL-LHC pile-up conditions and compared with existing Level-0 e/γ selection
methods. The CNN provides additional background rejection relative to the
baseline algorithms. Moreover, combining the CNN with the existing Level-0
e/γ selections yields the best overall performance for both single-electron
and di-electron trigger signatures.
| Do you plan to submit a 4-page extended abstract on OpenReview (only for Presentations/Posters)? | Maybe |
|---|