mlsanityData to 7 Oct 2026

Sparse Interaction Additive Networks via Feature Interaction Detection and Sparse Selection

arXiv 2209.09326 ↗First posted 19 Sep 2022cs.LG stat.MLcited by 28

Not on the map. Why?

Abstract (arXiv)

There is currently a large gap in performance between the statistically rigorous methods like linear regression or additive splines and the powerful deep methods using neural networks. Previous works attempting to close this gap have failed to fully investigate the exponentially growing number of feature combinations which deep networks consider automatically during training. In this work, we develop a tractable selection algorithm to efficiently identify the necessary feature combinations by leveraging techniques in feature interaction detection. Our proposed Sparse Interaction Additive Networks (SIAN) construct a bridge from these simple and interpretable models to fully connected neural networks. SIAN achieves competitive performance against state-of-the-art methods across multiple large-scale tabular datasets and consistently finds an optimal tradeoff between the modeling capacity of neural networks and the generalizability of simpler methods.

Key facts. Cited by 28 papers in the map (data through 7 Oct 2026). Builds on 18 of its 57 references that are in the map. First posted on arXiv 19 Sep 2022.

Builds on

18 of 57 references.

References · 18

  1. 2017Attention Is All You Needcited by 59,932
  2. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  3. 2017A Unified Approach to Interpreting Model Predictionscited by 6,488
  4. 2016XGBoost: A Scalable Tree Boosting Systemcited by 5,120
  5. 2017Axiomatic Attribution for Deep Networkscited by 3,350
  6. 2020Neural Additive Models: Interpretable Machine Learning with Neural Netscited by 318
  7. 2019InterpretML: A Unified Framework for Machine Learning Interpretabilitycited by 243
  8. 2019The Shapley Taylor Interaction Indexcited by 143
  9. 2020Explaining Explanations: Axiomatic Feature Interactions for Deep Networkscited by 103
  10. 2021NODE-GAM: Neural Generalized Additive Model for Interpretable Deep Learningcited by 88
  11. 2020GAMI-Net: An Explainable Neural Network based on Generalized Additive Models with Structured Interactionscited by 82
  12. 2020How does this interaction affect me? Interpretable attribution for feature interactionscited by 73
6 more
  1. 2020How Interpretable and Trustworthy are GAMs?cited by 44
  2. 2022Sparse Neural Additive Model: Interpretable Deep Learning with Feature Selection via Group Sparsitycited by 25
  3. 2016Additive Approximations in High Dimensional Nonparametric Regression via the SALSAcited by 8
  4. 2021Partially Interpretable Estimators (PIE): Black-Box-Refined Interpretable Machine Learningcited by 3
  5. 2016Learning Sparse Additive Models with Interactions in High Dimensionscited by 2
  6. 2021Creating Powerful and Interpretable Models with Regression Networkscited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 5; 2024: 9; 2025: 10; 2026: 4.

28 papers.

Sort:

  1. 2023TEMPO: Prompt-based Generative Pre-trained Transformer for Time Series Forecastingcited by 170
  2. 2023Neural Additive Models for Location Scale and Shape: A Framework for Interpretable Neural Regression Beyond the Meancited by 18
  3. 2025A Comprehensive Survey on Self-Interpretable Neural Networkscited by 18
  4. 2024Arithmetic Feature Interaction Is Necessary for Deep Tabular Learningcited by 8
  5. 2023Regionally Additive Models: Explainable-by-design models minimizing feature interactionscited by 3
  6. 2025Additive Models Explained: A Computational Complexity Approachcited by 3
  7. 2023Towards Hybrid-grained Feature Interaction Selection for Deep Sparse Networkcited by 2
  8. 2024Transparent Networks for Multivariate Time Seriescited by 2
  9. 2024A Complete Decomposition of KL Error using Refined Information and Mode Interaction Selectioncited by 2
  10. 2025Beyond Black-Box Predictions: Identifying Marginal Feature Effects in Tabular Transformer Networkscited by 2
  11. 2025Challenges in interpretability of additive modelscited by 2
  12. 2025Tree-like Pairwise Interaction Networkscited by 2
16 more
  1. 2025Higher-Order Causal Structure Learning with Additive Modelscited by 2
  2. 2024Neural Additive Image Model: Interpretation through Interpolationcited by 1
  3. 2024Discovering Car-following Dynamics from Trajectory Data through Deep Learningcited by 1
  4. 2024GAMformer: Bridging Tabular Foundation Models and Interpretable Machine Learningcited by 1
  5. 2024Generalized Sparse Additive Model with Unknown Link Functioncited by 1
  6. 2026Selecting Feature Interactions for Generalized Additive Models by Distilling Foundation Modelscited by 1
  7. 2023SPARLING: Learning Latent Representations with Extremely Sparse Activationscited by 0
  8. 2024Fast and interpretable Support Vector Classification based on the truncated ANOVA decompositioncited by 0
  9. 2024PE: A Poincare Explanation Method for Fast Text Hierarchy Generationcited by 0
  10. 2025HRR: Hierarchical Retrospection Refinement for Generated Image Detectioncited by 0
  11. 2025Multiplicative-Additive Constrained Models:Toward Joint Visualization of Interactive and Independent Effectscited by 0
  12. 2025Sparse Deep Additive Model with Interactions: Enhancing Interpretability and Predictabilitycited by 0
  13. 2025Additive Large Language Models for Semi-Structured Textcited by 0
  14. 2026EviNAM: Intelligibility and Uncertainty via Evidential Neural Additive Modelscited by 0
  15. 2026Neural Additive Experts: Context-Gated Experts for Controllable Model Additivitycited by 0
  16. 2026TetrisCNN for interpretable detection of phases of matter from experimental quantum simulator datacited by 0