Notifications
Clear all
Topic starter
19/08/2026 3:58 pm
TL;DR: Shapley values offer a fair, axiomatic way to attribute model predictions to input features, and Fiddler’s explainer shows why they became a standard tool for ML explainability and adverse action reasoning. The practical constraint is that feature absence must be approximated and the computation can explode to 2^n model calls, so attribution quality depends on sampling design and uncertainty handling.
NHIMG editorial — based on content published by Fiddler: AI Explained Video Series: What are Shapley Values?
Questions worth separating out
Q: How should teams decide whether Shapley values are suitable for a model explanation workflow?
A: Use Shapley values when you need a principled, axiomatic attribution method and can define feature absence in a defensible way.
Q: Why do Shapley explanations become unreliable in some tabular models?
A: Tabular models often require synthetic assumptions about what it means for a feature to be absent.
Q: What should practitioners measure before trusting a Shapley attribution result?
A: They should measure stability across samples, sensitivity to the missing-feature baseline, and the variance introduced by approximation.
Practitioner guidance
- Define the counterfactual rule for missing features Document exactly how your team represents feature absence in each model class, including baseline substitution or sampling logic for tabular data.
- Record uncertainty around attribution outputs Require confidence intervals, sampling variance, or stability checks for any Shapley-based explanation used in a decision workflow.
- Match the explainability method to the data structure Use direct feature removal only where absence is meaningful and observable, such as text inputs.
What's in the full article
Fiddler's full blog covers the implementation detail this post intentionally leaves for the source:
- Worked examples of how Shapley marginal contributions are computed across permutations
- Further detail on baseline substitution versus sampling-based feature absence for tabular models
- Discussion of computational shortcuts and why approximation introduces uncertainty
- The preprint guidance referenced by the author on choosing distributions for absent features
👉 Read Fiddler's explainer on what Shapley values are and how they support ML explainability →
Shapley values in ML explainability: where do the limits show up?
Explore further
19/08/2026 4:36 pm
Shapley values remain one of the clearest examples of explainability done with formal discipline, but they are not a governance shortcut. The method is valuable because it ties attribution to explicit axioms rather than intuition. Yet the quality of the explanation still depends on the missing-feature model, which means governance teams should treat it as a structured estimate rather than a definitive truth statement.
A question worth separating out:
Q: How do governance teams evaluate model explanations for regulated decisions?
A: They should review whether the explanation method is traceable, repeatable, and aligned to the data type and decision context. For regulated workflows, the explanation process should be documented well enough that an auditor or reviewer can see how the attribution was generated and where its limits begin.
👉 Read our full editorial: Shapley values expose the trade-off between fairness and compute
