What do you get when you marry #diffusion models with #Petri-net models? A better algorithm for trace #recovery. The paper “Efficient process embedding for diffusion-based trace recovery”, co-authored by Maximilian Matyash, Avigdor Gal, and Arik Senderovich (published in the #Process Science Journal and available here: https://rdcu.be/fmdLj) is an extended version of the best-paper award winner of #ICPM’2025. The paper describes how an encoding of a process model using #GNN in a diffusion model can assist in recovering a deterministic log from a stochastic one.