# Considerations When Learning Additive Explanations for Black-Box Models

arXiv: https://arxiv.org/abs/1801.08640 · HTML: https://mlsanity.com/paper/1801.08640/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 26 Jan 2018.

## Built on by (most cited first, 10 of 10)

- [Towards the Dynamics of a DNN Learning Symbolic Interactions](https://mlsanity.com/paper/2407.19198/) (2024)
- [Unifying Feature-Based Explanations with Functional ANOVA and Cooperative Game Theory](https://mlsanity.com/paper/2412.17152/) (2024)
- [NeurCAM: Interpretable Neural Clustering via Additive Models](https://mlsanity.com/paper/2408.13361/) (2024)
- [Function Trees: Transparent Machine Learning](https://mlsanity.com/paper/2403.13141/) (2024)
- [T-Explainer: A Model-Agnostic Explainability Framework Based on Gradients](https://mlsanity.com/paper/2404.16495/) (2024)
- [Disentangling Visual Transformers: Patch-level Interpretability for Image Classification](https://mlsanity.com/paper/2502.17196/) (2025)
- [textttfindr: Transparent and Fair Credit Risk Decisions through Semi-Structured Regressions](https://mlsanity.com/paper/2608.24582/) (2026)
- [Deep Explainable Learning with Graph Based Data Assessing and Rule Reasoning](https://mlsanity.com/paper/2211.04693/) (2022)
- [midr: Learning from Black-Box Models by Maximum Interpretation Decomposition](https://mlsanity.com/paper/2506.08338/) (2025)
- [Restricted Receptive Fields for Face Verification](https://mlsanity.com/paper/2510.10753/) (2025)

## Abstract (arXiv)

Many methods to explain black-box models, whether local or global, are additive. In this paper, we study global additive explanations for non-additive models, focusing on four explanation methods: partial dependence, Shapley explanations adapted to a global setting, distilled additive explanations, and gradient-based explanations. We show that different explanation methods characterize non-additive components in a black-box model's prediction function in different ways. We use the concepts of main and total effects to anchor additive explanations, and quantitatively evaluate additive and non-additive explanations. Even though distilled explanations are generally the most accurate additive explanations, non-additive explanations such as tree explanations that explicitly model non-additive components tend to be even more accurate. Despite this, our user study showed that machine learning practitioners were better able to leverage additive explanations for various tasks. These considerations should be taken into account when considering which explanation to trust and use to explain black-box models.