Feature attribution methods are central to explainable artificial intelligence. Their assumptions are expressed in several mathematical languages: cooperative-game values, path integrals, gradient operators, perturbation distributions, and backpropagation rules. This survey proposes a common framework for local additive feature attribution, organizing Shapley, path-based, gradient/backpropagation, perturbation, and CAM-style methods around five specification choices:
- Value Function
- Reference
- Path
- Perturbation Distribution
- Conservation Rule
It then compares these methods through an axiom-by-method matrix and links common failure modes, including baseline sensitivity, off-manifold perturbations, sanity-check failures, adversarial manipulation, and method disagreement, to the assumptions that produce them. Finally, the survey proposes a ten-item reporting checklist for studies that utilize local additive attributions. The central message is that attribution results are meaningful only relative to the mathematical assumptions under which they are defined, and those assumptions should be reported.
Blogger's Review: This paper provides a systematic analysis in the field of explainable AI, emphasizing the mathematical foundations of attribution methods and their impact on results. By establishing a standardized framework, researchers can better understand and compare the strengths and weaknesses of different attribution methods, promoting more transparent research practices.