Single-model generality
In machine learning, single-model generality is the kind of generality where literally the same model (i.e. weights in a neural network, or whatever) can be used to solve many different problems. This contrasts to single-architecture generality.