A run captures values for an individual execution of an experimental step. Each run can have resources and property values assigned to it, and can be connected to runs on other steps of the same experiment.