Improved particle-flow event reconstruction with scalable neural networks for current and future particle detectors
Problem
Reconstructing particles efficiently and accurately in the highly granular detectors expected at the High-Luminosity LHC and Future Circular Collider is computationally challenging.
Why it matters
Fast, accurate event reconstruction is necessary for physics discovery at future colliders.
What the AI did
Supervised machine learning models (a graph neural network and a kernel-based transformer) were trained on full detector simulation to perform particle-flow reconstruction from tracks and calorimeter clusters.
Results
Best graph neural network improved jet transverse momentum resolution by up to 50% versus the rule-based algorithm; model portable across Nvidia, AMD and Habana hardware.
Impact
Demonstrated in full detector simulation that scalable ML reconstruction is accurate, fast and hardware-portable.
Record
| Venue | Communications Physics |
|---|---|
| Published | 2024 |
| Work type | article |
| Authors | Joosep Pata, Eric Wulff, Farouk Mokhtar, David Southwick, Mengke Zhang, Maria Girone, Javier Mauricio Duarte |
| Citations | 9 (OpenAlex, at import time) |
| References | 70 |
| Model or method | Graph neural network and kernel-based transformer |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1038/s42005-024-01599-5 |
| Open access | gold |
| Keywords | particle flow, graph neural network, transformer, jet resolution, collider detectors |
Importance rating
significant — thousands to millions affected
Abstract ties reconstruction to HL-LHC and FCC detectors and states it can significantly improve future collider measurements, underpinning a foundational scientific field.
More in Physics & Astronomy
| # | Paper | Year | Citations | Importance |
|---|---|---|---|---|
| 1 | Magnetic control of tokamak plasmas through deep reinforcement learning | 2022 | 739 |
7
|
| 2 | Solving the quantum many-body problem with artificial neural networks | 2017 | 2,205 |
7
|
| 3 | Learning high-accuracy error decoding for quantum processors | 2024 | 90 |
7
|
| 4 | Optimizing Quantum Error Correction Codes with Reinforcement Learning | 2019 | 158 |
7
|
| 5 | Machine-learning-assisted correction of correlated qubit errors in a topological code | 2018 | 141 |
7
|