B-jet Tagging Using a Hybrid Edge Convolution and Transformer Architecture

2026-03-22

AI Summary

This work proposes the Edge Convolution Transformer (ECT) model to enhance the discrimination of b-jets from c-jets and light-flavor jets in hadronic collisions. ECT uniquely integrates edge convolution with Transformer self-attention to jointly model track-level features—such as impact parameters and momentum significance—and jet-level observables, including vertex and kinematic variables. This architecture preserves local geometric structure while capturing long-range dependencies. Evaluated on ATLAS simulation data, ECT achieves an AUC of 0.9333, substantially outperforming ParticleNet (0.8904) and a pure Transformer baseline (0.9216), with notable gains in rejecting the most challenging c-jet background. Moreover, its per-jet inference latency remains below 0.060 milliseconds, satisfying the real-time triggering requirements of the LHC.

Technology Category

Machine Learning: Learning on the Edge & Model CompressionComputer Vision: Diffusion Models for VisionReasoning under Uncertainty: Relational Probabilistic Models

Application Category

Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphsEconomics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAISemantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Jet flavor tagging plays an important role in precise Standard Model measurement enabling the extraction of mass dependence in jet-quark interaction and quark-gluon plasma (QGP) interactions. They also enable inferring the nature of particles produced in high-energy particle collisions that contain heavy quarks. The classification of bottom jets is vital for exploring new Physics scenarios in proton-proton collisions. In this research, we present a hybrid deep learning architecture that integrates edge convolutions with transformer self-attention mechanisms, into one single architecture called the Edge Convolution Transformer (ECT) model for bottom-quark jet tagging. ECT processes track-level features (impact parameters, momentum, and their significances) alongside jet-level observables (vertex information and kinematics) to achieve state-of-the-art performance. The study utilizes the ATLAS simulation dataset. We demonstrate that ECT achieves 0.9333 AUC for b-jet versus combined charm and light jet discrimination, surpassing ParticleNet (0.8904 AUC) and the pure transformer baseline (0.9216 AUC). The model maintains inference latency below 0.060 ms per jet on modern GPUs, meeting the stringent requirements for real-time event selection at the LHC. Our results demonstrate that hybrid architectures combining local and global features offer superior performance for challenging jet classification tasks. The proposed architecture achieves good results in b-jet tagging, particularly excelling in charm jet rejection (the most challenging task), while maintaining competitive light-jet discrimination comparable to pure transformer models.

Problem

• b-jet tagging
• jet flavor classification
• heavy quark identification
• charm rejection
• particle physics

Innovation

• Edge Convolution
• Transformer
• b-jet tagging
• hybrid architecture
• jet flavor tagging