Searcharxiv⌕ Search

arXiv subjects

Anru R. Zhang

Publications and source records attributed to Anru R. Zhang.

At least 19 recordsLinked to original sources
arXiv2026

Bias-Corrected Data Synthesis for Imbalanced Learning

Class imbalance complicates probabilistic classification because standard training objectives emphasize majority-class performance. Synthetic oversampling can reduce imbalance, but discrepancies between the synthetic and target minority distributions may bias the fitted classifier, especially because synthetic samples depend on the observed data. We propose a bias-correction procedure that estimates the generator-induced loss discrepancy from a held-out subset of majority observations and transfers this correction to the minority class under a uniform bias-transfer condition. We establish finite-sample bounds for bias transfer and for the excess balanced risk of the resulting empirical risk minimizer, and characterize a regime in which SMOTE induces non-negligible loss bias. The framework can also be implemented in imbalanced multi-task learning and propensity-score estimation, with details provided in the Supplementary Material. Real data analyses show that the correction is most useful when synthetic distortion is appreciable.

stat.ML↗
arXiv2026

Distilled Continuous Diffusion Language Models Can Write Code in Few Steps---or One

Language generation is almost universally treated as a sequential process: autoregressive models emit one token at a time, while diffusion language models replace token-level seriality with a long trajectory of iterative refinement. In this work, we introduce PlaidQ, a 0.7B continuous diffusion language model for code generation, and show that its trajectory can be aggressively distilled into only a few denoising steps---or even one, enabling efficient code generation. PlaidQ repurposes a pretrained autoregressive model as a bidirectional denoiser over continuous token embeddings. We distill PlaidQ with distribution matching for few-step generation and paired-trajectory supervision for one-step generation. At matched model scale, PlaidQ is competitive with discrete diffusion language models on code generation. Distillation then shifts the quality--compute frontier: a 16-step student reaches 31.78 and 40.49 pass@10 on HumanEval and MBPP+, surpassing the same PlaidQ teacher sampled for 512 steps. At the extreme, paired-trajectory distillation achieves 7.07 pass@1 on HumanEval with a single denoising step, producing functionally correct programs. Together, these results establish continuous diffusion as a viable path to few-step and one-step code generation. Broadly, continuous diffusion is not merely another representation for language: it provides an interface through which language models can inherit the acceleration and distillation machinery of continuous diffusion modeling. Training and inference code and model checkpoints are available at https://github.com/pengzhangzhi/plaidq.

cs.LG↗
arXiv2026

From Reverse Detection--Estimation Gaps to Computational Lower Bounds for Norm Approximation

We develop a general reduction scheme that converts reverse detection--estimation gaps into computational lower bounds for norm approximation. If an efficiently computable estimator has error well below a conjectured computational detection threshold, then a sufficiently accurate norm approximator applied to this estimator would yield an efficient test, transferring average-case detection hardness to worst-case approximation hardness. The ratio between the detection and estimation scales becomes a lower bound on the achievable approximation factor. We apply the reduction in three settings. Using a previously established reverse gap for high-order cumulant tensors, we obtain a conditional lower bound for polynomial-time approximation of the order-$d$ tensor spectral norm. We then establish new reverse gaps for the $k$-sparse matrix spectral norm and a normalized $k$-sparse matrix cut norm under planted-clique hardness. The resulting lower bounds are of order $\sqrt{k}$ up to a logarithmic factor, matching the simple deterministic $\sqrt{k}$ certificates, whereas previous hardness results for the sparse principal component objective rule out only constant factors. Finally, under a model-specific low-degree conjecture, a new reverse gap forces polynomially growing approximation factors for the order-$d$ tensor cut norm when $d\ge3$. These results identify reverse detection--estimation gaps as a systematic source of computational lower bounds for norm approximation.

math.ST↗
arXiv2026

Functional Post-Clustering Selective Inference with Applications to EHR Data Analysis

In electronic health records (EHR) analysis, clustering patients according to patterns in their data is crucial for uncovering new subtypes of diseases. Existing medical literature often relies on classical hypothesis testing methods to test for differences in means between these clusters. Due to selection bias induced by clustering algorithms, the implementation of these classical methods on post-clustering data often leads to an inflated type-I error. In this paper, we introduce a new statistical approach that adjusts for this bias when analyzing data collected over time. Our method extends classical selective inference methods for cross-sectional data to longitudinal data. We provide theoretical guarantees for our approach with upper bounds on the selective type-I and type-II errors. We apply the method to simulated data and real-world Acute Kidney Injury (AKI) EHR datasets, thereby illustrating the advantages of our approach.

stat.ME↗
arXiv2026

Smooth Flow Matching for Synthesizing Functional Data

Functional data, i.e., random functions observed over a continuous domain, are increasingly available in areas such as biomedical research, health informatics, and epidemiology. However, effective statistical analysis for functional data is often hindered by challenges such as privacy constraints, sparse and irregular sampling, infinite-dimensionality, and non-Gaussian structures. To address these challenges, we introduce a novel framework named Smooth Flow Matching (SFM), tailored for generative modeling of functional data that enables statistical analysis without exposing sensitive real data. Under a copula framework, SFM constructs a semiparametric smooth flow to generate infinite-dimensional functional data, free of Gaussianity and low-rank assumptions. It is computationally efficient, handles irregular observations, and guarantees the smoothness of the generated functions, offering a practical and flexible solution in scenarios where existing deep generative methods are not applicable. Through extensive simulation studies, we demonstrate the advantages of SFM in terms of both synthetic data quality and computational efficiency. We then apply SFM to generate clinical trajectory data from the MIMIC-IV patient electronic health records (EHR) longitudinal database. Our analysis showcases the ability of SFM to produce high-quality surrogate data for downstream tasks, highlighting its potential to boost the utility of EHR data for clinical applications.

stat.ML↗
arXiv2026

Bandable Cumulant Tensors: Optimal Estimation and Applications in Non-Gaussian Data Modeling

Higher-order cumulants capture the non-Gaussian dependence that covariance misses, but they are hard to use in high dimensions. An order-$d$ cumulant tensor has $p^d$ entries, and the plug-in sample cumulant is generally not even rate-optimal under the tensor spectral norm. For ordered data, both difficulties admit one remedy: assuming that higher-order interactions decay away from the main tensor diagonal, we introduce a bandable cumulant class and a tapered sample cumulant estimator that computes only $O(pk^{d-1})$ local entries at bandwidth $k$ and never forms the full tensor. Under exponential-type tail conditions, we prove nonasymptotic spectral-norm bounds that separate tapering bias from stochastic error, and a minimax lower bound over the same class that matches the leading bias and stochastic terms of the upper bound; for sub-Gaussian observations, the tapered estimator attains the minimax rate whenever $n\gg(k+\log p)^{d-1}$ at the oracle bandwidth $k$, with the ambient dimension entering only through $\log p$. Localization also suppresses the higher-order fluctuations behind this suboptimality, so tapering plays a stronger role here than in bandable covariance estimation. The spectral-norm guarantee transfers directly to downstream tasks, yielding plug-in error bounds for cumulant Yule--Walker estimation in autoregressive models, minimum-distance estimation in moving-average models, and matched-filter source localization in sensor arrays. Simulations corroborate the theory, and real-data analyses of RR-interval, air-quality, and Neuropixels recordings illustrate the resulting stability gains.

stat.ME↗
arXiv2026

Blessing of dimension in Bayesian inference on covariance matrices

Bayesian factor analysis is routinely used for dimensionality reduction in modeling of high-dimensional covariance matrices. Factor analytic decompositions express the covariance as a sum of low-rank and diagonal matrices. In practice, Gibbs sampling algorithms are typically used for posterior computation, alternating between updating the latent factors, loadings, and residual variances. In this article, we exploit a blessing of dimensionality to develop a provably accurate posterior approximation for the covariance matrix that bypasses the need for Gibbs or other variants of Markov chain Monte Carlo sampling. Our proposed Factor Analysis with BLEssing of dimensionality (FABLE) approach relies on a first-stage singular value decomposition (SVD) to estimate the latent factors, and then defines a jointly conjugate prior for the loadings and residual variances. The accuracy of the resulting posterior approximation for the covariance improves with increasing samples as well as increasing dimensionality. We show that FABLE has excellent performance in high-dimensional covariance matrix estimation, including producing well-calibrated credible intervals, both theoretically and through simulation experiments. We also demonstrate the strength of our approach in terms of accurate inference and computational efficiency by applying it to a gene expression dataset.

stat.ME↗
arXiv2026

Toward Automated Detection of Documentation Inconsistencies in Electronic Health Records

Objective: To characterize the kinds of internal documentation inconsistencies a general-domain large language model (LLM) can surface from real-world discharge summaries, and to identify recurring failure modes that limit reliability at scale. Materials and Methods: We applied a two-stage LLM pipeline---open-ended candidate identification (Gemini 2.5 Pro) followed by context-grounded verification (Gemini 2.5 Flash)---to 3,000 randomly sampled MIMIC-IV-Note discharge summaries. A subset of the pipeline output was then reviewed manually by clinical experts. Results: Our pipeline surfaced 3,460 candidate inconsistencies, affecting 69.7% of admissions. Representative examples spanned demographics, allergies, procedures, diagnoses, laboratory, medications, and care-planning domains, with direct implications for clinical reasoning or patient safety. Expert review also revealed recurring failure modes that arise when verification requires temporal reasoning, evolving-diagnosis context, or knowledge of outpatient-prescribing conventions the model does not natively possess. Discussion: Detection is highly context-dependent: many flagged pairs require anchoring each statement to its source section and clinical domain, then assessing whether the conflict reflects a true contradiction or missing context. We propose a graded ontology spanning strict contradiction and ambiguity, with a schema characterizing each flagged case by category, section, domain, and inconsistency axis. Conclusion: This formative study establishes a methodological foundation and conceptual framework to guide subsequent validated, large-scale EHR-inconsistency analysis.

cs.CL↗
arXiv2026

A Provably Efficient Method for Tensor Ring Decomposition and Its Applications

We present the first deterministic, finite-step algorithm for exact tensor ring (TR) decomposition. Our method leverages blockwise simultaneous diagonalization to recover TR cores from a limited number of tensor observations, under a dimension condition requiring each mode size to be at least quadratic in the TR rank and under a genericity assumption on the cores, thereby providing both algebraic insight and practical efficiency. We extend the approach to the symmetric TR setting, where parameter complexity is significantly reduced and applications arise naturally in physics-based modeling and exchangeable data analysis. To handle noisy observations, we develop a robust recovery scheme that couples our initialization with alternating least squares, achieving faster convergence and improved accuracy compared to classical methods. As applications, we obtain new algorithms for questions in other domains where tensor ring decomposition is a key primitive, namely matrix product state tomography in quantum information and provable learning of pushforward distributions in the foundations of machine learning. These contributions advance the algorithmic foundations of TR decomposition and open new opportunities for scalable tensor network computation.

math.NA↗
arXiv2026

A functional tensor model for dynamic multilayer networks with common invariant subspaces and the RKHS estimation

Dynamic multilayer networks are frequently used to describe the structure and temporal evolution of multiple relationships among common entities, with applications in fields such as sociology, economics, and neuroscience. However, exploration of analytical methods for these complex data structures remains limited. We propose a functional tensor-based model for dynamic multilayer networks, with the key feature of capturing the shared structure among common vertices across all layers, while simultaneously accommodating smoothly varying temporal dynamics and layer-specific heterogeneity. The proposed model and its embeddings can be applied to various downstream network inference tasks, including dimensionality reduction, vertex community detection, analysis of network evolution periodicity, visualization of dynamic network evolution patterns, and evaluation of inter-layer similarity. We provide an estimation algorithm based on functional tensor Tucker decomposition and the reproducing kernel Hilbert space framework, with an effective initialization strategy to improve computational efficiency. The estimation procedure can be extended to address more generalized functional tensor problems, as well as to handle missing data or unaligned observations. We validate our method on simulated data and two real-world cases: the dynamic Citi Bike trip network and an international food trade dynamic multilayer network, with each layer corresponding to a different product.

stat.ME↗
arXiv2026

Panel Flow Matching: A Generative Approach to Learning Distributions of Longitudinal Data

Learning distributions of longitudinal data is central to tasks such as visualization, completion, classification, and synthetic data generation, but it remains statistically challenging because longitudinal observations are often irregular, sparse, and collected from only a limited number of subjects. To address this, we develop a novel generative framework, termed panel flow matching (PFM), for learning longitudinal distributions by pooling information across time via a continuous panel flow model. PFM combines a forward flow-matching step with a backward kernel-fitting step, yielding a flexible and data-adaptive approach for capturing complex distributional structures. We apply PFM to estimate panel densities, namely the cross-sectional densities of longitudinal data, and establish statistical guarantees under irregular and sparse sampling designs. Under this, PFM naturally supports tasks including longitudinal completion, synthetic data generation, and classification, without requiring a preliminary dimension-reduction step to handle data irregularity. Extensive simulations demonstrate that PFM outperforms existing methods across these tasks. We further apply PFM to a vaginal microbiome longitudinal dataset from 188 pregnancies labeled as term or preterm, where it improves classification accuracy and reveals time-varying distributional differences between the two groups.

stat.ME↗
arXiv2026

When Does Synthetic Data Augmentation Improve Score-Based Imbalanced Classification?

Synthetic data augmentation is widely used to mitigate class imbalance, but its theoretical effects on score-based classification remain poorly understood. This paper develops a framework for characterizing when synthetic minority augmentation can improve threshold-integrated and threshold-optimized metrics, including AUROC, AUPRC, best-threshold balanced accuracy, and best-threshold \(\F_1\) score. We separate the effect of augmentation into two components: a change in effective class weighting and a discrepancy between the synthetic and true minority distributions. Under well-specified score models, the raw estimator already targets the likelihood-ratio ordering, which is population-optimal for the metrics considered. Consequently, augmentation cannot provide a fundamental population-level improvement beyond possible finite-sample variance reduction, and may introduce additional bias through synthetic distributional error. We further establish minimax lower bounds showing that the raw estimator already achieves the optimal metric-regret rate in the well-specified regime. Under misspecification, however, augmentation can play a qualitatively different role: by changing the effective class balance, it can alter the restricted-class projection and correct ranking errors induced by the raw imbalanced objective. We provide explicit improvement bounds quantifying the roles of approximation error, finite-sample estimation error, and synthetic distributional error. Simulation studies corroborate the theory, demonstrating limited gains under well-specification and nontrivial but nonmonotone improvements under misspecification.

stat.ML↗
arXiv2026

Optimal Estimation of Discrete Multiview Distributions under Heteroskedastic Multinomial Sampling

Multiview latent-variable models provide a fundamental framework for discrete data analysis, with applications to latent structure models, topic models, and mixtures of product distributions. In the discrete setting, the joint distribution of the observed views can be represented as a nonnegative low-rank tensor, which we call a multiview density tensor. We study the problem of estimating this tensor from multinomial count data. A key challenge is that multinomial sampling induces heteroskedastic and dependent noise, so the difficulty of estimation depends not only on the ambient dimensions and rank, but also on how the probability mass is distributed across different locations of sample space. We propose a general scaling framework for density tensor estimation under multinomial sampling. This framework leads to a spectral estimator for which we prove a Frobenius-norm upper bound that directly handles heteroskedasticity and negative dependence. For the original multiview model, we obtain fiber-mass-dependent Frobenius upper bounds and minimax lower bounds showing that this dependence is unavoidable. Under $\ell_1$ loss, we develop both oracle and feasible data-driven estimators based on the same scaling principle, establish minimax lower bounds, and show near-optimality for the oracle rule at fixed rank and for slice normalization under bounded slice-to-fiber imbalance. Simulations support the theory and demonstrate the robustness of the proposed methods.

stat.ME↗
arXiv2026

Coupling Models for One-Step Discrete Generation

Generative modeling over discrete structures underpins applications across deep learning, from biological sequence design and code generation to large language models, yet generation often remains sequential, relying on autoregressive decoding or iterative refinement. In this work, we introduce Coupling Models(Coupling Models), a one-step discrete generative model that learns a direct coupling between discrete sequences and Gaussian latents. Unlike recent distillation methods that compress a pretrained multi-step sampler into a few steps, Coupling Model trains a purpose-built decoder to invert this coupling and generate samples in a single step. The model also avoids complex continuous flows over the simplex and hand-specified data-to-noise couplings. Empirically,Coupling Model improves the strongest one-step baselines in each domain: it reduces LM1B text-generation perplexity by 33% at its lowest-perplexity operating point, Fly Brain enhancer-design FBD by 18%, and MNIST-Binary FID by 46%. These results suggest that effective one-step discrete generation depends strongly on how data and noise are coupled before decoding. Code is available at https://github.com/pengzhangzhi/Coupling-Models.

cs.LG↗
arXiv2026

Don't Retrain, Align: Adapting Autoregressive LMs to Diffusion LMs via Representation Alignment

Diffusion language models (DLMs) have recently demonstrated capabilities that complement standard autoregressive (AR) models, particularly in non-sequential generation and bidirectional editing. Although recent work has shown that pretrained autoregressive checkpoints can be converted into diffusion language models, existing recipes primarily transfer parameters through continued denoising training with objective- and attention-level modifications. We instead ask whether the internal representation geometry learned by next-token prediction can be explicitly preserved during AR-to-DLM conversion. We hypothesize that much of the semantic structure learned by AR pretraining can transfer across generation orders, and thus DLM training should be viewed as relearning the decoding path rather than relearning language representations. To investigate this, we introduce REPR-ALIGN, a representation alignment objective that adapts a bidirectional masked diffusion model to reuse representations from a pretrained AR model of identical architecture. Concretely, we align the hidden states of the DLM to the frozen AR model at every layer using cosine similarity, while optimizing the standard masked denoising objective. This simple alignment, with no adapters and no architectural changes beyond the attention mask, yields up to 4x training acceleration in our setting and is particularly effective in low-data regimes. Our results suggest that linguistic representations can transfer across generation order, and that representation alignment provides a simple and effective technique for training diffusion language models. Code is available at https://github.com/pengzhangzhi/Open-dLLM.

cs.LG↗
arXiv2026

Detection Is Harder Than Estimation in Certain Regimes: Inference for Moment and Cumulant Tensors

We study estimation and detection of high-order moment and cumulant tensors from $n$ i.i.d.\ observations of a $p$-dimensional random vector, with performance measured in tensor spectral norm. Under sub-Gaussianity, we show that the minimax rate for estimating the order-$d$ moment and cumulant tensors is $\sqrt{p/n}\wedge 1$. In contrast to covariance estimation, the sample moment tensor is generally not rate-optimal for $d\ge 3$, and we construct an estimator that attains the minimax rate up to logarithmic factors. On the computational side, we study testing whether the $d$-th order cumulant tensor vanishes after whitening. Using the low-degree polynomial framework, we provide evidence that detection is computationally hard when $n\ll p^{d/2}$. At the same time, we identify a regime in which an efficiently computable estimator achieves error smaller than the separation at which low-degree tests can reliably distinguish the null from the alternative. This reveals an unusual reverse detection--estimation gap: computationally efficient detection can be harder than computationally efficient estimation. The underlying reason is that the relevant loss, tensor spectral norm, is itself NP-hard to compute, creating a new form of computational--statistical gap.

math.ST↗
arXiv2026

Synthetic Augmentation in Imbalanced Learning: When It Helps, When It Hurts, and How Much to Add

Imbalanced classification often causes standard training procedures to prioritize the majority class and perform poorly on rare but important cases. A classic and widely used remedy is to augment the minority class with synthetic samples, but two basic questions remain under-resolved: when does synthetic augmentation actually help, and how many synthetic samples should be generated? We develop a unified statistical framework for synthetic augmentation in imbalanced learning, studying models trained on imbalanced data augmented with synthetic minority samples. Our theory shows that synthetic data is not always beneficial. In a "local symmetry" regime, imbalance is not the dominant source of error, so adding synthetic samples cannot improve learning rates and can even degrade performance by amplifying generator mismatch. When augmentation can help ("local asymmetry"), the optimal synthetic size depends on generator accuracy and on whether the generator's residual mismatch is directionally aligned with the intrinsic majority-minority shift. This structure can make the best synthetic size deviate from naive full balancing. Practically, we recommend Validation-Tuned Synthetic Size (VTSS): select the synthetic size by minimizing balanced validation loss over a range centered near the fully balanced baseline, while allowing meaningful departures. Extensive simulations and real data analysis further support our findings.

stat.ML↗
arXiv2026

Planner Aware Path Learning in Diffusion Language Models Training

Diffusion language models have emerged as a powerful alternative to autoregressive models, enabling fast inference through more flexible and parallel generation paths. This flexibility of sampling is unlocked by new engineered sampling strategies, or planners, that select more favorable generation paths by iteratively planning - versus uniformly at random - where to denoise along the sequence. However, by modifying the reverse paths via planning, planners create an irrevocable mismatch between the uniformly random denoising paths assumed during training and planning-based inference. In this paper, we systematically investigate the mismatch of discrete diffusion training and inference under planning and theoretically prove that the standard discrete diffusion training evidence lower bound (ELBO) does not accurately describe a denoiser that uses a non-uniform planner. To address this gap, we derive a new planned evidence lower bound (P-ELBO) that incorporates planner-based reverse dynamics directly into the training objective. Using the P-ELBO, we introduce Planner Aware Path Learning (PAPL), a novel training scheme that aligns training and inference under a planned denoiser. PAPL is implemented as a simple yet effective modification to the standard masked discrete diffusion loss, making it widely applicable and easy to adopt. Empirically, we show PAPL delivers consistent gains across domains, including a 40% relative improvement in protein sequences, improved text generation with up to a 4x relative MAUVE gain, and 23% relative improvement in code generation HumanEval pass@10. Code is available at github.com/pengzhangzhi/PAPL .

cs.LG↗