Global Interpretability

The ability to understand the overall logic and behaviour of an entire AI model, including which features drive predictions across all inputs.

Understanding the big picture of how an AI works overall; what patterns it learned, what it generally considers important. Like understanding a company's overall hiring philosophy.

Understanding a model's overall behaviour is essential for governance oversight. Global interpretability enables risk managers and auditors to assess whether an AI system's logic aligns with organisational policies and values.