Responsible AI

Interpretability & Explainability

Interpretability is understanding how a model works; explainability is communicating why a particular output happened. The need depends on risk, debugging needs, user trust, and regulation.
  • Simple models can be inherently interpretable
  • Global explanations are not local explanations
  • Local explanation tools are approximations
  • Explanations must fit the audience
  • LLM explanations may be post-hoc narratives
  • Evidence beats eloquence
Explanation types
TypeExampleUseful for
Model-intrinsicSmall tree path, linear coefficientsDebugging and audit
Global post-hocFeature importanceUnderstanding broad behavior
Local post-hocSHAP/LIME for one predictionCase review
Evidence-basedSource passages / tool tracesGrounded user explanation
CounterfactualWhat would need to change?Recourse and decision support
Sources
  • An Introduction to Statistical Learning with PythonInterpretable models
  • scikit-learn User GuideInspection
  • Artificial Intelligence: A Modern ApproachAI risks and explainability