用生成模型加速大型强子对撞机中子探测器仿真，更快更准。
Fine-tuned Normalizing Flows for ALICE Zero Degree Calorimeter Fast Simulation
- 基于归一化流的生成模型，通过迁移学习和分步微调适配不同粒子类型。
- 在多个物理指标上优于基线，水氏距离达1.61±0.02，提升仿真精度。
- 引入新评估指标，更适合捕捉粒子响应的条件依赖与波动特性，适合高能物理仿真研究者。
在大型强子对撞机（LHC）中模拟ALICE零度量能器（ZDC）中子探测器响应计算成本高昂，需复杂的蒙特卡洛链。本文开发了一种生成代理模型，聚焦于归一化流（Normalizing Flows, NFs）。通过迁移学习，在全不平衡数据集上预训练，并采用两种渐进解冻策略，为不同粒子类型（γ, n, Λ, K_S^0, Σ^+）微调专用模型。由于标准指标如水氏距离忽略条件结构，我们提出改进指标：条件加权平均绝对误差、离散比和雅各布共激活误差，更准确捕捉物理相关的输入-输出依赖与响应变异性。所提出的集成模型在所有指标上均优于基线，水氏距离为1.61±0.02。本工作提供了一个通用的基于归一化流的LHC探测器仿真框架，融合了归一化流、条件微调与物理驱动的评估方法。
原文摘要 · Abstract (English)
Simulating the ALICE Zero Degree Calorimeter (ZDC) neutron detector responses at the LHC is computationally expensive, requiring complex Monte Carlo chains. We develop a generative surrogate, focusing on Normalizing Flows (NFs). Through transfer learning, we pre-train on the full imbalanced dataset and fine-tune specialized models for different particle types ($γ$, $n$, $Λ$, $K_S^0$, $Σ^+$) using two gradual-unfreezing schemes. As standard ZDC metrics like Wasserstein distance overlook conditional structure, we introduce refined metrics: conditional weighted MAE, dispersion ratio, and Jaccard co-activation error, that better capture physics-relevant input-output dependencies and response variability. Our ensemble of fine-tuned models achieves a Wasserstein distance of $1.61 \pm 0.02$, outperforming baselines across all metrics. This work provides a generalizable NF-based framework for LHC detector simulation, combining NFs, conditional fine-tuning, and physics-motivated evaluation.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。