B BSTabDiff
NeurIPS 2026 · Main Conference Poster Previously · ICLR 2026 DeLTa Workshop

BSTabDiff

Block-Subunit Diffusion Priors for High-Dimensional Tabular Data Generation

Learn global dependence in a compact space of block subunits, then decode high-dimensional synthetic samples with realistic marginals, dependencies, and optional missingness.

Al Zadid Sultan Bin Habib · Md Younus Ahamed · Prashnna Kumar Gyawali · Gianfranco Doretto · Donald A. Adjeroh

West Virginia University · University of Utah

NeurIPS logo
18 Real datasets across multiple tabular regimes
42,728 Features in the Cell Cycle dataset
10 Distinct generator baselines across the evaluation
89.21% Extended nine-dataset mean TSTR AUC, no-permutation variant
Browse page sections

Paper Abstract

Abstract

High-Dimensional Low-Sample Size (HDLSS) tabular domains (e.g., omics) are characterized by \(n \ll m\), where \(n\) = number of samples, and \(m\) = number of features. Such domains often exhibit strong local correlation groups, sparse cross-group dependencies, heavy-tailed non-Gaussian marginals, heteroscedastic noise, and structured missingness, making direct density learning in \(\mathbb{R}^m\) ill-conditioned since \(n \ll m\). We propose BSTabDiff, a block-subunit generative framework that partitions the \(m\) observed features into \(M\) latent blocks (\(M \ll m\)) and generates each block via a shared low-dimensional subunit variable, concentrating global dependence learning in the compact block-latent space \(\mathbb{R}^M\) while decoding to the full feature space with copula-driven dependence, flexible per-feature marginals, and explicit missingness mechanisms. BSTabDiff can induce a data-driven feature ordering and block construction before latent partitioning, allowing the model to better align with local dependency structure when such organization is beneficial. BSTabDiff supports modern deep priors on block latents, including diffusion and normalizing flows, enabling stable synthesis and controllable benchmark generation in the HDLSS regime. Empirically, BSTabDiff produces more realistic and more stable high-dimensional synthetic data when compared with unstructured tabular generators on HDLSS data.

Motivation and Contributions

Why Block-Subunit Generation?

Conventional tabular generators often model relationships over all features at once. When there are only dozens or hundreds of samples but thousands of columns, learning dense dependence becomes ill-conditioned. BSTabDiff identifies groups of related features and learns how those groups interact through shared factors.

\[ X \in \mathbb{R}^{m}, \quad h=(h_1,\ldots,h_M)\in\mathbb{R}^{M}, \quad M\ll m,\quad n\ll m \] Full-dimensional observations, compact block-latent dependence.
Block-subunit model

Each related feature group shares one latent variable.

Deep latent prior

Diffusion or normalizing flows model cross-block dependencies.

Structured decoding

Copula variables, marginal transforms, and missingness.

Feature ordering

GO-BS estimates column order and contiguous block boundaries.

HDLSS-oriented analysis

Block factorization, identifiability, and sample efficiency.

Sustainable learning

Utility, fidelity, compute, energy, and carbon evaluation.

Architecture

End-to-End Generative Pipeline

BSTabDiff separates block-level dependence learning from full-dimensional feature emission. The architecture integrates optional class conditioning, diffusion/flow priors, missingness, inverse marginal decoding, and observed-feature permutation.

BSTabDiff block-subunit generative architecture
BSTabDiff generative architecture. Full-resolution figure ↗

Generation in Five Stages

  1. Optional class prior Draw a label \(Y \sim p(Y)\), or fix a class for conditional generation.
  2. Sample compact block latents Draw \(h \sim p_\theta(h\mid Y)\) from a diffusion or flow prior.
  3. Generate observation mask Construct \(\widetilde R \in \{0,1\}^{m}\) from a feature-specific missingness process.
  4. Decode each feature Use block-conditioned Gaussian copula variables and learned inverse marginal CDFs.
  5. Restore feature order Return generated features to their observed indexing when ordering has been applied.

Core Methodology

Explore the Mathematical Formulation

Select a stage to explore BSTabDiff's six-part generative formulation. The equations describe the methodology, while the interactive diagrams provide conceptual illustrations rather than outputs from a trained model.

Step 1 of 6

Latent Block Partition

Let canonical feature indices be divided into M disjoint blocks. Features in the same block share the latent subunit hₜ.

\[ \{1,\dots,m\} =\bigsqcup_{t=1}^{M}S_t, \qquad h\sim p_\theta(h\mid Y) \]

Feature partition and block-latent prior.

Instead of learning dependencies directly across all m features, BSTabDiff learns global dependence among M block latents.

Conceptual block illustration

A toy example with 12 features. Adjust M to change the partition.

Mathematical notation is rendered using MathJax. The interactive illustrations demonstrate the methodology conceptually.

Structure-Aware Learning

Graph-guided Ordering with Block Segmentation

GO-BS constructs a weighted feature-affinity graph and searches for a feature permutation and contiguous block boundaries that keep strongly dependent features together.

Optimization Objective

\[ \mathcal{J}_{\mathrm{GO-BS}}(\pi,b) = -A_{\mathrm{within}}(\pi,b) +\lambda A_{\mathrm{cross}}(\pi,b) +\gamma\Omega(b) \]

Within-block affinity is rewarded, cross-block affinity is penalized, and the block size regularizer \(\Omega(b)\) controls the resulting partition structure.

Interactive Feature Ordering Explorer

Eight toy features form two planted correlation groups (blue and peach). Select any two feature chips to swap their positions. Reordering updates the affinity matrix display and recalculates the within-block, cross-block, and objective scores.

Feature Order & Affinity Matrix

Darker squares mean higher affinity \(W_{ij}\). Blue dividers mark block boundaries. Diagonal cells are omitted.

Live Objective Breakdown

Within-block affinity ↑ —
Cross-block affinity ↓ —
Boundary regularizer (toy Ω) —
Number of boundaries —
Illustrative objective J (lower is better) —

Important: This is an educational example, not the trained GO-BS optimizer. Its boundary surrogate is \(\Omega_{\mathrm{toy}}(b)=M-1\); it does not claim that the paper uses this particular regularizer. Affinities and group memberships are illustrative.

Mathematical Components

Within-block Affinity (Reward)
\[ A_{\mathrm{within}}(\pi,b) = \sum_{t=1}^{M} \sum_{\substack{ i,j\in B_t(\pi,b)\\ j>i }} W_{ij} \]

Counts each unordered pair of features once when both features fall inside the same block.

Cross-block Affinity (Penalty)
\[ A_{\mathrm{cross}}(\pi,b) = \sum_{t=1}^{M-1} \sum_{t'=t+1}^{M} \sum_{i\in B_t(\pi,b)} \sum_{j\in B_{t'}(\pi,b)} W_{ij} \]

Counts each unordered pair of features once when they fall in different blocks.

Here, \(B_t(\pi,b)\) denotes the t-th contiguous feature block, \(W_{ij}\) is the affinity between features i and j, and \(\lambda\) and \(\gamma\) control the cross-block and boundary terms. The objective corresponds to Eq. 10 of the NeurIPS 2026 paper.

GO-BS

Sample-clustered graph ordering for coherent feature block construction.

GO-BS-FC

Feature-clustered variant offering improved compute–structure trade-offs in extreme dimensions.

Feature ordering can be optional. Where local feature structure is weak, ordering may add computation without improving utility. On the Cell Cycle dataset (42,728 features), GO-BS-FC achieved the strongest TSTR performance among the evaluated variants.

Theory and Interpretation

Why BSTabDiff Helps in HDLSS

BSTabDiff leverages compact block-latent representations and shared within-block structure to improve modeling in High-Dimensional, Low-Sample Size (HDLSS) settings. The theoretical analysis studies block identifiability and the benefits of signal aggregation under specific assumptions.

1. Compact Latent Modeling

Instead of learning global dependencies directly across all \(m\) observed features, BSTabDiff concentrates dependence learning in \(M\) block latents, where \(M \ll m\).

100× Fewer global dependence coordinates in the block-latent representation
Original feature space 10,000
Block-latent space 100
\[ X\in\mathbb{R}^{m}, \qquad h\in\mathbb{R}^{M}, \quad M\ll m \]

The ratio \(m/M\) represents the reduction in dependence-space dimensionality. It is not a guaranteed reduction in runtime, memory use, or parameter count.

2. Signal Aggregation

When features within a block share a class-dependent mean shift and have independent Gaussian noise, averaging them reduces conditional variance and improves signal-to-noise ratio.

Illustrative assumptions: \(\mu_1-\mu_0=1\) and \(\sigma=2\).

0.500 Conditional variance
2.00 Aggregated SNR
Single-feature SNR 0.25
Aggregated SNR 2.00
\[ \operatorname{Var} (\overline X_S\mid Y) = \frac{\sigma^2}{s} \] \[ \operatorname{SNR} (\overline X_S) = s\frac{(\mu_1-\mu_0)^2}{\sigma^2} \]

Simplified mean-shift Gaussian example corresponding to Lemma 3.5. Increasing \(s\) improves SNR linearly under the stated assumptions; correlated noise or heterogeneous feature signals can change this behavior.

Theoretical interpretation. The paper investigates identifiability of block structure up to block-label permutations under distinct block-dependence assumptions. The GO-BS optimization objective is NP-hard, motivating approximate solvers. These results and sample-efficiency interpretations are conditional on their stated assumptions rather than universal guarantees.

Experimental Evaluation

TSTR vs TRTR · 18 Datasets

TSTR = Train on Synthetic, Test on Real; TRTR = Train on Real, Test on Real. Bars show the highest reported mean BSTabDiff TSTR for each dataset alongside the real data reference. Use the controls to switch between Accuracy and ROC-AUC.

Evaluation Metric
Primary HDLSS · Accuracy
BSTabDiff TSTR Real TRTR
Figure: BSTabDiff synthetic data utility compared with training on real data. Each bar uses a fixed 0–100% scale. Variant labels identify the selected configuration. Accuracy and AUC are shown separately.

Dataset Dimensions

n = samples; m = features.

Dataset n m

Sources: BSTabDiff (NeurIPS 2026), Tables 2 and 4, and Appendix Tables A8.1, A18.1, and A23.3. Results show dataset-level means; standard deviations are provided in the paper. CIFAR-10 uses ResNet-50 image embeddings; Springleaf uses the processed subset. The best variant may differ across datasets.

Compute and Sustainability

Efficiency and Responsible Interpretation

Compact prior

Learn cross-block dependence in \(M\) latent variables, not all \(m\) features.

Fast HDLSS synthesis

Reported no-permutation runs typically finish in under 10 seconds on Vulcan (NVIDIA RTX A6000) hardware.

Energy-carbon analysis

The evaluation measures utility, energy, CO₂e, memory, and computational cost.

Emission fitting scales with observed sample-feature entries, while the main diffusion dependence-learning cost depends on the smaller latent dimension \(M\). GO-BS preprocessing may add significant computation in extreme feature spaces.

BSTabDiff is designed primarily for HDLSS data. Synthetic data should complement, rather than replace, real data validation and expert review. Fidelity and downstream utility remain dataset dependent, and the method does not imply formal differential privacy.

Open-Source Implementation

Installation and Reproducibility

BSTabDiff is available on PyPI and can be installed directly using pip. Source code, notebooks, smoke tests, and experimental instructions are available in the GitHub repository.

pip install bstabdiff

Publication History

Paper, Posters, and Resources

BSTabDiff was previously presented at the ICLR 2026 DeLTa Workshop and subsequently accepted as a NeurIPS 2026 main conference poster. Please cite the NeurIPS version for the main conference publication.

Research Support

Acknowledgements

U.S. National Science Foundation logo

National Science Foundation

This work was supported in part by the US National Science Foundation (NSF) under Awards 1920920, 2125872, 2223793, and 2318708.

NeurIPS 2026 logo

NeurIPS 2026 Review Process

We sincerely thank the anonymous NeurIPS 2026 reviewers for their valuable feedback and constructive suggestions. We are especially grateful to the anonymous Area Chair for their thoughtful insights and guidance, which significantly contributed to strengthening this work.

Computational Support

We also gratefully acknowledge Zaigham Abbas Randhawa, Pouya Iranmanesh, and Ghazaleh Mirzaee for their assistance with GPU clusters.

Citation

Cite BSTabDiff

If you use BSTabDiff or its published results, please cite our NeurIPS 2026 paper.

@inproceedings{habib2026bstabdiff,
  title     = {BSTabDiff: Block-Subunit Diffusion Priors for High-Dimensional Tabular Data Generation},
  author    = {Habib, Al Zadid Sultan Bin and Ahamed, Md Younus and Gyawali, Prashnna Kumar and Doretto, Gianfranco and Adjeroh, Donald A.},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year      = {2026}
}