Foundations

Interpretability and fairness interview questions

Sooner or later someone asks why the model made a particular decision, and an interviewer wants to know what you’d say. These questions test whether you know the difference between explaining a model overall and explaining one prediction, and what tools like SHAP can and can’t claim.

The fairness side asks where bias gets into data and models, which fairness metric fits a given situation, and why you usually can’t satisfy all of them at once. Good answers are honest about limits: showing what the model relied on is not the same as showing why the outcome happened.

Try one now

Standard

A stakeholder asks why the model declined a particular customer. The model is gradient boosting. What do you tell them?

Free, no sign-up. The interviewer asks follow-ups on your reasoning, then rates your answer.

What the 9 questions cover

  • Fairness metrics4
  • Global vs local explanations3
  • Sources of bias in data and models3
  • Feature importance2
  • SHAP values2
  • Explaining a model to stakeholders2
  • Partial dependence and ICE plots1

A question can cover more than one topic, so the counts add up to more than 9.

How the practice works

  1. The AI interviewer asks the question and follows up on your answer, the way an interviewer would.
  2. Stuck? Ask for a hint. Each one gives a little more away, so you can get unstuck without being handed the answer.
  3. Get a written debrief. Four ratings, each from Needs work to Interview-ready, tied to what you actually said, and what to study next.

More foundations questions

See all categories