The inherent value of SDGs in the model is captured implicitly in the AI model training. As we train the model, the neural network learns the patterns of value for different SDGs.
As such, different SDGs will yield different values, and differently for the values of the other attributes (standard, project activity, geography) as inferred from the data.