Explainable AI vs. Mechanistic Interpretability
NewBoth are about understanding why an AI made a decision, so 'interpretability' on a resume can mean either. Explainable AI (XAI) gives a practical explanation of a prediction — which factors pushed a loan decision one way — and is often required by regulators in banking, insurance and healthcare. Mechanistic interpretability is research that opens up a neural network to work out how it actually computes its answers, mostly done at AI labs. The first is applied data science; the second is frontier research.
How to tell them apart on a resume
Explainable AI (XAI)
SHAP, LIME, feature importance, model risk management, fairness audits, regulatory reporting — banks, insurers, healthcare, credit scoring.
Mechanistic interpretability
Sparse autoencoders, circuits, features, activation patching, TransformerLens, published research — AI labs and safety research groups.
The question that settles it
“Were you explaining individual predictions to the business or a regulator, or researching how the network works internally?”
Read the full definitions
Open the full tool for the other look-alike pairs, role profiles, and the JD decoder.