Neural Petri flows for chemical reactions

Key points

• 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.1
• We find the answer in the theory, where all semantics of a net share the firing form $m^\prime=m+Cσ$, locality, as enabling reads only the inputs of a transition, and the enabling rule, and we prove that conservation forces the firing form and that non-negativity forces the enabling rule on local rate laws.1
• On what we denote a valence net, atom mapping, reaction classification, and forward prediction become three tasks on one firing vector.1
• With electrons as tokens, the same token game predicts 90.5% of the elementary steps of FlowER first, ahead of the published baseline, and every top-1 prediction is a valid molecule without a filter.1

Sources (1)

• [1]Neural Petri flows for chemical reactions
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 6, 05:44 PM

  We introduce Neural Petri Flow, which learns this rate law, or a readout for classification, and hard-wires the rest as parameter-free layers.

  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.

Extractive summary: sentences quoted from the sources.