How do generative AI models help data scientists select the most appropriate model architecture for a data set?

By generating explanations about how the model made a particular prediction

By measuring the mutual information between features and the target variable

By pitting their two neural networks, generator and discriminator, against each other

By generating latent data representations that capture the underlying structure of the data