Despite extensive efforts to create fairer machine learning (ML) datasets, there remains a limited understanding of the practical aspects of dataset curation.

We consider two versions: sequential, where Bob observes Alice's cut

The remarkable capability of over-parameterised neural networks to generalise effectively has been explained by invoking a "simplicity bias": neural networks prevent overfitting by initially learning simple classifiers before progressing to

However, in the real world, the graph topology and features are inevitably evolving over time, e.g., the user connections and interests in social networks.