mlsanityData to 14 Oct 2026

Towards Interpretable Deep Neural Networks for Tabular Data

arXiv 2509.08617 ↗First posted 10 Sep 2025cs.LGcited by 1

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

Abstract (arXiv)

Tabular data is the foundation of many applications in fields such as finance and healthcare. Although DNNs tailored for tabular data achieve competitive predictive performance, they are blackboxes with little interpretability. We introduce XNNTab, a neural architecture that uses a sparse autoencoder (SAE) to learn a dictionary of monosemantic features within the latent space used for prediction. Using an automated method, we assign human-interpretable semantics to these features. This allows us to represent predictions as linear combinations of semantically meaningful components. Empirical evaluations demonstrate that XNNTab attains performance on par with or exceeding that of state-of-the-art, black-box neural models and classical machine learning approaches while being fully interpretable.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Sparse autoencoder training and evaluation, one of 13 directions in Mechanistic interpretability. Builds on 17 of its 19 references that are in the map. First posted on arXiv 10 Sep 2025.

Builds on

17 of 19 references.

References · 17

  1. 2016XGBoost: A Scalable Tree Boosting Systemcited by 5,166
  2. 2023Sparse Autoencoders Find Highly Interpretable Features in Language Modelscited by 1,420
  3. 2021Transformers in Vision: A Surveycited by 1,013
  4. 2021Revisiting Deep Learning Models for Tabular Datacited by 740
  5. 2019TabNet: Attentive Interpretable Tabular Learningcited by 648
  6. 2014OpenML: networked science in machine learningcited by 526
  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. 2024TabM: Advancing Tabular Deep Learning with Parameter-Efficient Ensemblingcited by 113
  10. 2021Transformer visualization via dictionary learning: contextualized embedding as a linear superposition of transformer factorscited by 108
  11. 2021A Comprehensive Taxonomy for Explainable Artificial Intelligence: A Systematic Survey of Surveys on Methods and Conceptscited by 91
  12. 2023TabR: Tabular Deep Learning Meets Nearest Neighbors in 2023cited by 74
5 more
  1. 2022T2G-Former: Organizing Tabular Features into Relation Graphs Promotes Heterogeneous Feature Interactioncited by 41
  2. 2023TANGOS: Regularizing Tabular Neural Networks through Gradient Orthogonalization and Specializationcited by 32
  3. 2023ProtoGate: Prototype-based Neural Networks with Global-to-local Feature Selection for Tabular Biomedical Datacited by 18
  4. 2024Binning as a Pretext Task: Improving Self-Supervised Learning in Tabular Domainscited by 12
  5. 2024InterpreTabNet: Distilling Predictive Signals from Tabular Data by Salient Feature Interpretationcited by 10

Built on by

Cited by year

Citing papers per year: 2025: 1.

1 paper.

  1. 2025XNNTab -- Interpretable Neural Networks for Tabular Data using Sparse Autoencoderscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024A Survey on Deep Tabular Learning
  2. 2025TabNSA: Native Sparse Attention for Efficient Tabular Data Learning
  3. 2024Mambular: A Sequential Model for Tabular Deep Learning
  4. 2026ReSS: Learning Reasoning Models for Tabular Data Prediction via Symbolic Scaffold
  5. 2026MaskTab: Scalable Masked Tabular Pretraining with Scaling Laws and Distillation for Industrial Classification
  6. 2024MambaTab: A Plug-and-Play Model for Learning Tabular Data
  7. 2026XGenBoost: Synthesizing Small and Large Tabular Datasets with XGBoost
  8. 2024TabDPT: Scaling Tabular Foundation Models on Real Data