Explain why gradient boosting fits residuals, and what that implies about how it fails.
Answer it out loud before you open anything. The value of the flags below is in comparing them to what you actually said — including whether you asked about the data before naming a model.
A pricing team runs a gradient-boosted model with 3,000 trees that was tuned last year. A colleague proposes "adding another 2,000 trees to improve it" and separately proposes averaging it with a random forest "because ensembles are always better". Both suggestions came from the same intuition that more trees are better.
What it is really testing
Whether the candidate understands the mechanism — each tree is fitted to the gradient of the loss with respect to the current prediction, which for squared error is the residual — well enough to predict behaviour: why more trees eventually overfit, why the learning rate and tree count trade off, and why a forest and a boosted model are different objects.