Title: QINNs: Quantum-Informed Neural Networks Show Chinese title

Title: QINNs：量子信息神经网络

Abstract: Classical deep neural networks can learn rich multi-particle correlations in collider data, but their inductive biases are rarely anchored in physics structure. We propose quantum-informed neural networks (QINNs), a general framework that brings quantum information concepts and quantum observables into purely classical models. While the framework is broad, in this paper, we study one concrete realisation that encodes each particle as a qubit and uses the Quantum Fisher Information Matrix (QFIM) as a compact, basis-independent summary of particle correlations. Using jet tagging as a case study, QFIMs act as lightweight embeddings in graph neural networks, increasing model expressivity and plasticity. The QFIM reveals distinct patterns for QCD and hadronic top jets that align with physical expectations. Thus, QINNs offer a practical, interpretable, and scalable route to quantum-informed analyses, that is, tomography, of particle collisions, particularly by enhancing well-established deep learning approaches. Show Chinese abstract

Abstract: 经典深度神经网络可以学习对撞机数据中的丰富多粒子相关性，但它们的归纳偏差很少与物理结构相联系。 我们提出量子启发神经网络（QINNs），这是一个通用框架，将量子信息概念和量子可观测量引入纯经典模型中。 虽然该框架范围广泛，但在本文中，我们研究了一个具体的实现，该实现将每个粒子编码为一个量子比特，并使用量子费舍尔信息矩阵（QFIM）作为粒子相关性的紧凑且与基无关的总结。 以喷注分类为例，QFIM在图神经网络中作为轻量级嵌入，提高了模型的表达能力和可塑性。 QFIM揭示了QCD和强子顶喷注的不同模式，这些模式与物理预期一致。 因此，QINNs提供了一种实用、可解释且可扩展的量子启发分析途径，即粒子碰撞的层析成像，特别是通过增强已建立的深度学习方法。