Question: When a model is being trained, optimization and generalization are correlated when 1. the model is performing better on validation data than on training data. 2. the model is performing better on training data than on validation data. 3. the models is overfitted. 4. the model is underfitted.