Scalable do-Shapley Explanations with Estimand-Agnostic Causal Inference

Álvaro Parafita, Tomas Garriga, Axel Brando, Francisco J. Cazorla

Keywords: Shapley, Explainability, Causality, Attribution, SCM, Modeling

TL;DR: Scalable do-SHAP with Structural Causal Models and a novel algorithm to detect reducible coalitions

Abstract:

Among explainability techniques, SHAP stands out as one of the most popular, but often overlooks the causal structure of the problem. While do-SHAP uses interventional causal queries, its reliance on estimands hinders scalability. To address this problem, we propose employing estimand-agnostic Causal Inference, which allows for the estimation of any identifiable query with a single model, making
do-SHAP feasible on arbitrarily complex graphs. We also develop a novel algorithm to significantly accelerate its computation at a negligible cost with a marked improvement in computational speed, as well as a method to explain inaccessible Data Generating Processes. We validate our approach on two real-world datasets, highlighting its potential in obtaining reliable explanations.

Primary Area: interpretability and explainable AI

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 6290