I am trying to wrap my head around how/why email name is used as a feature in the einstein engagement scoring models. If the model inputs are the individuals in the account subscriber list, does this mean there is a feature for every email send name, then marked 1 or 0 if that user opened that email? If that is the case, what if a specific email was never even sent to that individual? I would assume then it would impact the prediction for that user, even though it has no true impact on their behavior (since they never recieved it, which is drastically different from not opening it).