When do explicit penalties improve flow matching?

Lea Kunkel, Alberto González-Sanz, Sophie Langer

Flow matching can learn structure directly from data, raising the question of when additional constraints improve generation. We study penalties that encode such constraints, with…