arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2605.21789hep-excs.AI
收藏

基于补丁层次注意力的高效粒子喷注标记变压器

Patch Hierarchical Attention Transformer for Efficient Particle Jet Tagging

University of Illinois Chicago(伊利诺伊大学香槟分校);University of California San Diego(加州大学圣地亚哥分校);California Institute of Technology(加州理工学院);Fermi National Accelerator Laboratory(费米国家加速器实验室)

Aaron Wang, Zihan Zhao, Alan Xia, Chang Sun, Abhijith Gandrakota, Jennifer Ngadiuba, Richard Cavanaugh, Javier Duarte

AI总结:

本文提出了一种结合物理启发的几何信息传递模块和基于补丁的层次注意力机制的Patch Hierarchical Attention Transformer (PHAT-JeT),以在有限资源下实现高效的粒子喷注标记,从而在四个基准测试中取得最佳性能。

摘要:

实时喷注标记对于在大型强子对撞机的高通量探测器中识别短寿命粒子衰变至关重要,其中负责决定存储哪些碰撞事件的实时触发系统对延迟和准确性提出了严格要求。尽管变换器架构在计算不受限制时能够实现最高的喷注标记准确性,但其二次自注意力成本使得在触发预算内进行推理变得受限。现有的高效变体虽然降低了计算成本,但会阻碍分类性能。为了解决这一限制,我们引入了Patch Hierarchical Attention Transformer (PHAT-JeT),它结合了两个机制:一个受物理启发的几何信息传递模块,用于编码局部探测器平面结构,以及一个基于补丁的层次注意力方案,该方案在小粒子组内计算精确的注意力,同时通过轻量级补丁-标记通信保持全局上下文。在受限预算内,PHAT-JeT在四个基准测试(hls4ml、JetClass、Top Tagging 和 Quark-Gluon)中实现了所有资源受限喷注标记模型中的最佳准确性和背景拒绝率。我们的代码可在 https://github.com/aaronw5/PHAT-JeT 上获得。

英文摘要:

Real-time jet tagging is critical for identifying short-lived particle decays in the high-throughput detectors of the Large Hadron Collider, where real-time trigger systems responsible for deciding which collision events to store impose strict latency and accuracy constraints. While transformer architectures achieve the highest jet tagging accuracy when compute is unconstrained, their quadratic self-attention cost makes inference restrictive on trigger budget. Existing efficient variants reduce the computational cost, but hinder the classification performance. To address this limitation, we introduce the Patch Hierarchical Attention Transformer (PHAT-JeT), which combines two mechanisms: a physics-inspired geometric message-passing module that encodes local detector-plane structure, and a hierarchical patch-based attention scheme that computes exact attention within small particle groups while preserving global context through lightweight patch-token communication. Within a restricted budget, PHAT-JeT achieves state-of-the-art accuracy and background rejection among all resource-constrained jet tagging models on four benchmarks (\textsc{hls4ml}, JetClass, Top Tagging, and Quark--Gluon). Our code is available at https://github.com/aaronw5/PHAT-JeT.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑