What does “Generalization” mean?

When a model memorizes training data and performs poorly on unseen inputs.

Training approaches that let a model keep learning from new data without forgetting prior knowledge.

Techniques that create modified training examples to improve model generalization.

How clearly an AI system's logic, data sources, and limitations are documented and understandable.

How well a model's confidence scores match actual correctness probabilities.

A held-out dataset used to measure model quality after training.