[论文] Neural Petri flows for chemical reactions

论文概要

研究领域: ML 作者: Jose Eduardo Escrig Molina, Daniel Probst 发布时间: 2026-10-06 arXiv: 2610.08750

中文摘要

Petri网已被用于描述化学反应等化学过程。它们很好地映射到化学：库所是原子之间的键和每个原子的自由化合价，token是键序的单位，变迁形成或断裂键，守恒量是原子的化合价预算，使能规则是化合价规则。基于Petri网构建的化学反应学习模型或使用网作为消息传递支架的神经网络并不保证这些语义。在这里，我们问：什么架构对于其权重的每个值仍然是一个Petri网？我们在理论中找到了答案——网的所有语义共享触发形式m'=m+Cσ、局部性（使能只读取变迁的输入）和使能规则。我们证明了守恒强制触发形式，非负性在局部速率定律上强制使能规则。这留下了速率定律的自由——每个变迁触发的倾向。我们引入神经Petri流（NPF），学习这个速率定律或用于分类的读出，并将其余部分硬连线为无参数层。在化合价网上，原子映射、反应分类和前向预测成为同一触发向量上的三个任务。无需训练，最小触发向量映射了88.8%的策划Golden集（RXNMapper为85.6%），以及EnzymeMap的88.7%（对比77.9%）的酶促反应。在USPTO-480K上，NPF预测了87.7%的产物（1%训练子集上为67.4%）。ECREACT的EC编号在第三层级预测了90.2%的反应，领先最佳已发表方法5.6个百分点。

原文摘要

Petri nets have been used to describe chemical processes such as reactions. They map well to chemistry: Places are the bonds between atoms and the free valence of each atom, a token is a unit of bond order, a transition forms or breaks a bond, the conserved quantities are the valence budgets of the atoms, and the enabling rule is the valence rule. These semantics are not guaranteed by learned models of reactions or neural networks that are built on Petri nets that use the net as a scaffold for message passing. Here, we ask what architecture remains a Petri net for every value of its weights. We find the answer in the theory, where all semantics of a net share the firing form m'=m+Cσ, locality, as enabling reads only the inputs of a transition, and the enabling rule, and we prove that conse...

#论文 #arXiv #ML #小凯