mlsanityData to 14 Oct 2026

XNNTab -- Interpretable Neural Networks for Tabular Data using Sparse Autoencoders

arXiv 2512.13442 ↗First posted 15 Dec 2025cs.LGcited by 0

On the map: Mechanistic interpretability → Sparse autoencoder training and evaluation

Abstract (arXiv)

In data-driven applications relying on tabular data, where interpretability is key, machine learning models such as decision trees and linear regression are applied. Although neural networks can provide higher predictive performance, they are not used because of their blackbox nature. In this work, we present XNNTab, a neural architecture that combines the expressiveness of neural networks and interpretability. XNNTab first learns highly non-linear feature representations, which are decomposed into monosemantic features using a sparse autoencoder (SAE). These features are then assigned human-interpretable concepts, making the overall model prediction intrinsically interpretable. XNNTab outperforms interpretable predictive models, and achieves comparable performance to its non-interpretable counterparts.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Sparse autoencoder training and evaluation, one of 13 directions in Mechanistic interpretability. Builds on 20 of its 27 references that are in the map. First posted on arXiv 15 Dec 2025.

Builds on

20 of 27 references.

Closest ancestors · 1

Same direction, newest first

  1. 2025Towards Interpretable Deep Neural Networks for Tabular Data10 Sep

Other references · 19

  1. 2016XGBoost: A Scalable Tree Boosting Systemcited by 5,166
  2. 2019Optuna: A Next-generation Hyperparameter Optimization Frameworkcited by 2,501
  3. 2023Sparse Autoencoders Find Highly Interpretable Features in Language Modelscited by 1,420
  4. 2021Transformers in Vision: A Surveycited by 1,013
  5. 2021Revisiting Deep Learning Models for Tabular Datacited by 740
  6. 2019TabNet: Attentive Interpretable Tabular Learningcited by 648
  7. 2020TabTransformer: Tabular Data Modeling Using Contextual Embeddingscited by 409
  8. 2021SAINT: Improved Neural Networks for Tabular Data via Row Attention and Contrastive Pre-Trainingcited by 319
  9. 2019Neural Oblivious Decision Ensembles for Deep Learning on Tabular Datacited by 203
  10. 2024TabM: Advancing Tabular Deep Learning with Parameter-Efficient Ensemblingcited by 113
  11. 2021Transformer visualization via dictionary learning: contextualized embedding as a linear superposition of transformer factorscited by 108
  12. 2021A Comprehensive Taxonomy for Explainable Artificial Intelligence: A Systematic Survey of Surveys on Methods and Conceptscited by 91
7 more
  1. 2023TabR: Tabular Deep Learning Meets Nearest Neighbors in 2023cited by 74
  2. 2022T2G-Former: Organizing Tabular Features into Relation Graphs Promotes Heterogeneous Feature Interactioncited by 41
  3. 2023TANGOS: Regularizing Tabular Neural Networks through Gradient Orthogonalization and Specializationcited by 32
  4. 2023ProtoGate: Prototype-based Neural Networks with Global-to-local Feature Selection for Tabular Biomedical Datacited by 18
  5. 2024Binning as a Pretext Task: Improving Self-Supervised Learning in Tabular Domainscited by 12
  6. 2024InterpreTabNet: Distilling Predictive Signals from Tabular Data by Salient Feature Interpretationcited by 10
  7. 2025This looks like what? Challenges and Future Research Directions for Part-Prototype Modelscited by 3

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026XGenBoost: Synthesizing Small and Large Tabular Datasets with XGBoost
  2. 2024Mambular: A Sequential Model for Tabular Deep Learning
  3. 2024MambaTab: A Plug-and-Play Model for Learning Tabular Data
  4. 2026MaskTab: Scalable Masked Tabular Pretraining with Scaling Laws and Distillation for Industrial Classification
  5. 2025TabNSA: Native Sparse Attention for Efficient Tabular Data Learning
  6. 2024A Survey on Deep Tabular Learning
  7. 2022TabPFN: A Transformer That Solves Small Tabular Classification Problems in a Second
  8. 2025Early Stopping Tabular In-Context Learning