Papers

Mutational signature analysis has emerged as a powerful method for uncovering the underlying biological processes driving cancer development. However, the signature extraction process, typically performed using non-negative matrix factorization (NMF), often lacks reliability and clinical applicability. To address these limitations, several solutions have been introduced, including the use of neural networks to achieve more accurate estimates and probabilistic methods to better capture natural variation in the data. In this work, we introduce a Variational Autoencoder for Mutational Signatures (VAE-MS), a novel model that leverages both an asymmetric architecture and probabilistic methods for the extraction of mutational signatures. VAE-MS is compared to with three state-of-the-art models for mutational signature extraction: SigProfilerExtractor, the NMF-based gold standard; MUSE-XAE, an autoencoder that employs an asymmetric design without probabilistic components; and SigneR, a Bayesian NMF model, to illustrate the strength in combining a nonlinear extraction with a probabilistic model. In the ability to reconstruct input data and generalize to unseen data, models with probabilistic components (VAE-MS, SigneR) dramatically outperformed models without (SigProfilerExtractor, MUSE-XAE). The NMF-baed models (SigneR, SigProfilerExtractor) had the most accurate reconstructions in simulated data, while VAE-MS reconstructed more accurately on real cancer data. Upon evaluating the ability to extract signatures consistently, no model exhibited a clear advantage over the others. Software for VAE-MS is available at https://github.com/CLINDA-AAU/VAE-MS.

Hyperdimensional Computing (HDC) offers a computationally efficient paradigm for neuromorphic learning. Yet, it lacks rigorous uncertainty quantification, leading to open decision boundaries and, consequently, vulnerability to outliers, adversarial perturbations, and out-of-distribution inputs. To address these limitations, we introduce ConformalHDC, a unified framework that combines the statistical guarantees of conformal prediction with the computational efficiency of HDC. For this framework, we propose two complementary variations. First, the set-valued formulation provides finite-sample, distribution-free coverage guarantees. Using carefully designed conformity scores, it forms enclosed decision boundaries that improve robustness to non-conforming inputs. Second, the point-valued formulation leverages the same conformity scores to produce a single prediction when desired, potentially improving accuracy over traditional HDC by accounting for class interactions. We demonstrate the broad applicability of the proposed framework through evaluations on multiple real-world datasets. In particular, we apply our method to the challenging problem of decoding non-spatial stimulus information from the spiking activity of hippocampal neurons recorded as subjects performed a sequence memory task. Our results show that ConformalHDC not only accurately decodes the stimulus information represented in the neural activity data, but also provides rigorous uncertainty estimates and correctly abstains when presented with data from other behavioral states. Overall, these capabilities position the framework as a reliable, uncertainty-aware foundation for neuromorphic computing.

In this paper, we study last-iterate convergence of learning algorithms in bilinear saddle-point problems, a preferable notion of convergence that captures the day-to-day behavior of learning dynamics. We focus on the challenging setting where players select actions from compact convex sets and receive only bandit feedback. Our main contribution is the design of an uncoupled learning algorithm that guarantees last-iterate convergence to the Nash equilibrium with high probability. We establish a convergence rate of $\tilde{O}(T^{-1/4})$ up to polynomial factors in problem parameters. Crucially, our proposed algorithm is computationally efficient, requiring only an efficient linear optimization oracle over the players' compact action sets. The algorithm is obtained by combining techniques from experimental design and the classic Follow-The-Regularized-Leader (FTRL) framework, with a carefully chosen regularizer function tailored to the geometry of the action set of each learner.

We consider the problem of sampling from a posterior distribution arising in Bayesian inverse problems in science, engineering, and imaging. Our method belongs to the family of independence Metropolis-Hastings (IMH) sampling algorithms, which are common in Bayesian inference. Relying on the existence of an approximate posterior distribution that is cheaper to sample from but may have significant bias, we introduce Proximal-IMH, a scheme that removes this bias by correcting samples from the approximate posterior through an auxiliary optimization problem. This yields a local adjustment that trades off adherence to the exact model against stability around the approximate reference point. For idealized settings, we prove that the proximal correction tightens the match between approximate and exact posteriors, thereby improving acceptance rates and mixing. The method applies to both linear and nonlinear input-output operators and is particularly suitable for inverse problems where exact posterior sampling is too expensive. We present numerical experiments including multimodal and data-driven priors with nonlinear input-output operators. The results show that Proximal-IMH reliably outperforms existing IMH variants.

In this work, we consider causal inference in various high-dimensional treatment settings, including for single multi-valued treatments and vector treatments with binary or continuous components, when the number of treatments can be comparable to or even larger than the number of observations. These settings bring unique challenges: first, the treatment effects of interest are represented by a high-dimensional vector rather than a scalar; second, positivity violations are often unavoidable; and third, estimation can be based on a smaller effective sample size. We first discuss fundamental limits of estimating effects here, showing that consistent estimation is impossible without further assumptions. We go on to propose novel doubly robust estimators for mean potential outcomes of a high-dimensional single multi-valued treatment. We analyze the proposed estimators under sparsity assumptions, giving finite-sample risk bounds and showing that consistent estimation is possible under these conditions. Moreover, we derive minimax lower bounds in a sparse and structure-agnostic model to characterize optimal rates of convergence and show our risk bounds are unimprovable. We then generalize our proposed estimators as a sparse pseudo-outcome regression framework with constrained regression estimators and error guarantees under sparsity, allowing estimation of generic functionals and different types of high-dimensional treatments. We apply the framework to derive estimators of the mean potential outcomes for high-dimensional vector treatments. Finally, we illustrate the proposed methods through a simulation and an empirical application.

Sample overlap is a common issue in evidence synthesis in the field of medical research, particularly when integrating findings from observational studies utilizing existing databases such as registries. Due to the general inaccessibility of unique identifiers for each observation, addressing sample overlap has been a complex problem, potentially biasing evidence synthesis outcomes and undermining their credibility. We developed a method to construct indicators for the degree of sample overlap in evidence synthesis of studies based on existing data. Our method is rooted in set theory and is based on the coding of the ranges of several well selected sample characteristics, offers a practical solution by focusing on making inference based on sample characteristics rather than on individual participant data. Useful information, such as the overlap-free sample set with the largest sample size in an evidence synthesis, can be derived from this method. We applied our model to several real-world evidence syntheses, demonstrating its effectiveness and flexibility. Our findings highlight the growing importance of addressing sample overlap in evidence synthesis, especially with the increasing relevance of secondary use of data, an area currently under-explored in research.

Gaussian process (GP) surrogates are the default tool for emulating expensive computer experiments, but cubic cost, stationarity assumptions, and Gaussian predictive distributions limit their reach. We propose Generative Bayesian Computation (GBC) via Implicit Quantile Networks (IQNs) as a surrogate framework that targets all three limitations. GBC learns the full conditional quantile function from input–output pairs; at test time, a single forward pass per quantile level produces draws from the predictive distribution. Across fourteen benchmarks we compare GBC to four GP-based methods. GBC improves CRPS by 11–26% on piecewise jump-process benchmarks, by 14% on a ten-dimensional Friedman function, and scales linearly to 90,000 training points where dense-covariance GPs are infeasible. A boundary-augmented variant matches or outperforms Modular Jump GPs on two-dimensional jump datasets (up to 46% CRPS improvement). In active learning, a randomized-prior IQN ensemble achieves nearly three times lower RMSE than deep GP active learning on Rocket LGBB. Overall, GBC records a favorable point estimate in 12 of 14 comparisons. GPs retain an edge on smooth surfaces where their smoothness prior provides effective regularization.

An index of an effective number of variables (ENV) is introduced for model selection in nested models. This is the case, for instance, when we have to decide the order of a polynomial function or the number of bases in a nonlinear regression, choose the number of clusters in a clustering problem, or the number of features in a variable selection application (to name few examples). It is inspired by the idea of the maximum area under the curve (AUC). The interpretation of the ENV index is identical to the effective sample size (ESS) indices concerning a set of samples. The ENV index improves {drawbacks of} the elbow detectors described in the literature and introduces different confidence measures of the proposed solution. These novel measures can be also employed jointly with the use of different information criteria, such as the well-known AIC and BIC, or any other model selection procedures. Comparisons with classical and recent schemes are provided in different experiments involving real datasets. Related Matlab code is given.

We study the problem of efficiently producing, in an online fashion, generative models of scalar, multiclass, and vector-valued outcomes that cannot be falsified on the basis of the observed data and a pre-specified collection of computational tests. Our contributions are twofold. First, we expand on connections between online high-dimensional multicalibration with respect to an RKHS and recent advances in expected variational inequality problems, enabling efficient algorithms for the former. We then apply this algorithmic machinery to the problem of outcome indistinguishability. Our procedure, Defensive Generation, is the first to efficiently produce online outcome indistinguishable generative models of non-Bernoulli outcomes that are unfalsifiable with respect to infinite classes of tests, including those that examine higher-order moments of the generated distributions. Furthermore, our method runs in near-linear time in the number of samples and achieves the optimal, vanishing T^{-1/2} rate for generation error.

Recently a new experimental approach, the hybrid experimental design (HED), was introduced to enable investigators to answer scientific questions about building behavioral interventions in which human-delivered and digital components are integrated and adapted on multiple timescales: slow (e.g., every few weeks) and fast (e.g., every few hours), respectively. An increasingly common HED involves the integration of the sequential, multiple assignment, randomized trial (SMART) with the micro-randomized trial (MRT), allowing investigators to answer scientific questions about potential synergistic effects of digital and human-delivered interventions. Approaches to formalize these questions in terms of causal estimands and associated data analytic methods are limited. In this paper, we formally define and assess these synergistic effects in hybrid SMART-MRTs on both proximal and distal outcomes. Practical utility is shown through the analysis of M-Bridge, a hybrid SMART-MRT aimed at reducing binge drinking among first-year college students.

Perturbed Utility Models (PUMs) represent discrete choice as utility maximization under convex perturbation or regularization, encompassing a broad class of choice models, including Multinomial Logit (MNL) and models that permit zero choice probabilities. Because, for general perturbations, maximum likelihood may be non-convex, and the negative log-likelihood is infinite when a sparse model assigns zero probability to an observed choice, we specialize the Fenchel–Young criterion to PUM estimation. With a fixed perturbation and utilities linear in parameters, the empirical objective is globally convex and remains finite at sparse predictions. We establish probability-level Fisher consistency, parameter identification, consistency, and asymptotic normality under explicit regularity conditions. We also develop computational methods for estimation, using structured probability evaluation for additive and tree-structured perturbations and a primal-dual algorithm for general non-separable perturbations. We then introduce Parametric Basis Estimation (PBE), a bi-level method that jointly learns utility parameters and a tree-structured perturbation through convex combinations of leaf-level and nest-level basis functions. Under explicit regularity conditions, we establish joint identification and consistency of the resulting estimator when the bi-level problem is solved globally. We also develop an algorithm with stationarity guarantees under stated smoothness conditions. Synthetic experiments examine structural recovery, normalization, and sensitivity to misspecification. On Swissmetro, repeated respondent-grouped nested cross-validation shows that both additive and tree-structured PBE achieve statistically significant reductions in held-out Brier score relative to MNL.

Traditional health authority approval for oncology drugs is based on a clinical benefit endpoint, or a valid surrogate. In 1992 the FDA created the Accelerated Approval pathway to allow for earlier approval of therapies in serious conditions with an unmet medical need. This is accomplished typically by granting accelerated approval based on a surrogate endpoint that can be measured earlier than a traditional approval endpoint. Minimal residual disease (MRD) is a sensitive measure of residual cancer cells in hematology oncology after treatment, and is increasingly considered as a secondary or exploratory endpoint due to its prognostic potential for traditional clinical trial endpoints such as progression-free survival (PFS) and overall survival (OS). This work aims to evaluate MRD's surrogacy potential across several hematologic cancer indications while keeping the focus on follicular lymphoma (FL), using data from published studies. We examine individual-level and trial-level correlations extracted from previously published studies to elucidate the potential role of MRD in accelerating the drug approval process in hematology oncology trials.

Given a black-box AI system and a task, at what confidence level can a practitioner trust the system's output? We answer with a reliability level – a single number per system-task pair, derived from self-consistency sampling and conformal calibration, that serves as a black-box deployment gate with exact, finite-sample, distribution-free guarantees. Self-consistency sampling reduces uncertainty exponentially; conformal calibration guarantees correctness within 1/(n+1) of the target level, regardless of the system's errors – made transparently visible through larger answer sets for harder questions. Weaker models earn lower reliability levels (not accuracy – see Definition 2.4): GPT-4.1 earns 94.6% on GSM8K and 96.8% on TruthfulQA, while GPT-4.1-nano earns 89.8% on GSM8K and 66.5% on MMLU. We validate across five benchmarks, five models from three families, and both synthetic and real data. Conditional coverage on solvable items exceeds 0.93 across all configurations; sequential stopping reduces API costs by around 50%.

This paper studies the means-testing problem under weakly correlated Normal setups. Although quite common in genomic applications, test procedures having exact FWER control under such dependence structures are nonexistent. We explore the asymptotic behaviors of the classical Bonferroni (when adjusted suitably) and the Sidak procedure; and show that both of these control FWER at the desired level exactly as the number of hypotheses approaches infinity. We derive analogous limiting results on the generalized family-wise error rate and power. Simulation studies depict the asymptotic exactness of the procedures empirically.

Bayesian inference for inverse problems involves computing expectations under posterior distributions–e.g., posterior means, variances, or predictive quantities–typically via Monte Carlo (MC) estimation. When the quantity of interest varies significantly under the posterior, accurate estimates demand many samples–a cost often prohibitive for partial differential equation-constrained problems. To address this challenge, we introduce conditional neural control variates, a modular method that learns amortized control variates from joint model-data samples to reduce the variance of MC estimators. To scale to high-dimensional problems, we leverage Stein's identity to design an architecture based on an ensemble of hierarchical coupling layers with tractable Jacobian trace computation. Training requires: (i) samples from the joint distribution of unknown parameters and observed data; and (ii) the posterior score function, which can be computed from physics-based likelihood evaluations, neural operator surrogates, or learned generative models such as conditional normalizing flows. Once trained, the control variates generalize across observations without retraining. We validate our approach on stylized and partial differential equation-constrained Darcy flow inverse problems, outperforming classical Stein control variates and achieving substantial variance reduction, even when the analytical score is replaced by a learned surrogate.

Based on the algorithm Informed Importance Tempering (IIT) proposed by Li et al. (2023) we propose an algorithm that uses an adaptive bounded balancing function. We argue why implementing parallel tempering where each replica uses a rejection free MCMC algorithm can be inefficient in high dimensional spaces and show how the proposed adaptive algorithm can overcome these computational inefficiencies. We present two equivalent versions of the adaptive algorithm (A-IIT and SS-IIT) and establish that both have the same limiting distribution, making either suitable for use within a parallel tempering framework. To evaluate performance, we benchmark the adaptive algorithm against several MCMC methods: IIT, Rejection free Metropolis-Hastings (RF-MH) and RF-MH using a multiplicity list. Simulation results demonstrate that Adaptive IIT identifies high-probability states more efficiently than these competing algorithms in high-dimensional binary spaces with multiple modes.

Representation learning has been essential for graph machine learning tasks such as link prediction, community detection, and network visualization. Despite recent advances in achieving high performance on these downstream tasks, little progress has been made toward self-explainable models. Understanding the patterns behind predictions is equally important, motivating recent interest in explainable machine learning. In this paper, we present GraphHull, an explainable generative model that represents networks using two levels of convex hulls. At the global level, the vertices of a convex hull are treated as archetypes, each corresponding to a pure community in the network. At the local level, each community is refined by a prototypical hull whose vertices act as representative profiles, capturing community-specific variation. This two-level construction yields clear multi-scale explanations: a node's position relative to global archetypes and its local prototypes directly accounts for its edges. The geometry is well-behaved by design, while local hulls are kept disjoint by construction. To further encourage diversity and stability, we place principled priors, including determinantal point processes, and fit the model under MAP estimation with scalable subsampling. Experiments on real networks demonstrate the ability of GraphHull to recover multi-level community structure and to achieve competitive or superior performance in link prediction and community detection, while naturally providing interpretable predictions.

Choi and Yuan (2025) propose a novel approach to applying matrix completion to the problem of estimating causal effects in panel data. The key insight is that even in the presence of structured patterns of missing data – i.e. selection into treatment – matrix completion can be effective if the number of treated observations is small relative to the number of control observations. We applaud the authors for their insightful and interesting paper. We discuss this proposal from two complementary perspectives. First, we situate their proposal as an example of a "split-apply-combine" strategy that underlies many modern panel data estimators, including difference-in-differences and synthetic control approaches. Second, we discuss the issue of the statistical "last mile problem" – the gap between theory and practice – and offer suggestions on how to partially address it. We conclude by considering the challenges of estimating the impacts of public policies using panel data and apply the approach to a study on the effect of right to carry laws on violent crime.

Neural networks are trained by optimizing multi-dimensional sets of fitting parameters on non-convex loss landscapes. Low-loss regions of the landscapes correspond to the parameter sets that perform well on the training data. A key issue in machine learning is the performance of trained neural networks on previously unseen test data. Here, we investigate neural network training by stochastic gradient descent (SGD) - a non-convex global optimization algorithm which relies only on the gradient of the objective function. We contrast SGD solutions with those obtained via a non-stochastic quasi-Newton method, which utilizes curvature information to determine step direction and Golden Section Search to choose step size. We use several computational tools to investigate neural network parameters obtained by these two optimization methods, including kernel Principal Component Analysis and a novel, general-purpose algorithm for finding low-height paths between pairs of points on loss or energy landscapes, FourierPathFinder. We find that the choice of the optimizer profoundly affects the nature of the resulting solutions. SGD solutions tend to be separated by lower barriers than quasi-Newton solutions, even if both sets of solutions are regularized by early stopping to ensure adequate performance on test data. When allowed to fit extensively on the training data, quasi-Newton solutions occupy deeper minima on the loss landscapes that are not reached by SGD. These solutions are less generalizable to the test data however. Overall, SGD explores smooth basins of attraction, while quasi-Newton optimization is capable of finding deeper, more isolated minima that are more spread out in the parameter space. Our findings help understand both the topography of the loss landscapes and the fundamental role of landscape exploration strategies in creating robust, transferrable neural network models.

Hamiltonian Monte Carlo (HMC) is a state of the art method for sampling from distributions with differentiable densities, but can converge slowly when applied to challenging multimodal problems. Running HMC with a time varying Hamiltonian, in order to interpolate from an initial tractable distribution to the target of interest, can address this problem. In conjunction with a weighting scheme to eliminate bias, this can be viewed as a special case of Sequential Monte Carlo (SMC) sampling \cite{doucet2001introduction}. However, this approach can be inefficient, since it requires slow change between the initial and final distribution. Inspired by \cite{sels2017minimizing}, where a learned counterdiabatic term added to the Hamiltonian allows for efficient quantum state preparation, we propose Counterdiabatic Hamiltonian Monte Carlo (CHMC), which can be viewed as an SMC sampler with a more efficient kernel. We establish its relationship to recent proposals for accelerating gradient-based sampling with learned drift terms, and demonstrate on simple benchmark problems.

We present Group Orthogonalized Policy Optimization (GOPO), a new alignment algorithm for large language models derived from the geometry of Hilbert function spaces. Instead of optimizing on the probability simplex and inheriting the exponential curvature of Kullback-Leibler divergence, GOPO lifts alignment into the Hilbert space L2(pi_k) of square-integrable functions with respect to the reference policy. Within this space, the simplex constraint reduces to a linear orthogonality condition <v, 1> = 0, defining a codimension-one subspace H0. Minimizing distance to an unconstrained target u_star yields the work-dissipation functional J(v) = <g, v> - (mu / 2) ||v||^2, whose maximizer follows directly from the Hilbert projection theorem. Enforcing the boundary v >= -1 produces a bounded Hilbert projection that induces exact sparsity, assigning zero probability to catastrophically poor actions through a closed-form threshold. To connect this functional theory with practice, GOPO projects from infinite-dimensional L2(pi_k) to a finite empirical subspace induced by group sampling. Because group-normalized advantages sum to zero, the Lagrange multiplier enforcing probability conservation vanishes exactly, reducing the constrained projection to an unconstrained empirical loss. The resulting objective has constant Hessian curvature mu I, non-saturating linear gradients, and an intrinsic dead-zone mechanism without heuristic clipping. Experiments on mathematical reasoning benchmarks show that GOPO achieves competitive generalization while maintaining stable gradient dynamics and entropy preservation in regimes where clipping-based methods plateau.

In multivariate longitudinal studies, associations between outcomes often exhibit time-varying and individual level heterogeneity, motivating the modeling of correlations as an explicit function of time and covariates. However, most existing methods for correlation analysis fail to simultaneously capture the time-varying and covariate-dependent effects. We propose a Time-Varying and Covariate-Dependent (TiVAC) correlation model that jointly allows covariate effects on correlation to change flexibly and smoothly across time. TiVAC employs a bivariate Gaussian model where the covariate-dependent correlations are modeled semiparametrically using penalized splines. We develop a penalized maximum likelihood-based Newton-Raphson algorithm, and inference on time-varying effects is provided through simultaneous confidence bands. Simulation studies show that TiVAC consistently outperforms existing methods in accurately estimating correlations across a wide range of settings, including binary and continuous covariates, sparse to dense observation schedules, and across diverse correlation trajectory patterns. We apply TiVAC to a psychiatric case study of 291 bipolar I patients, modeling the time-varying correlation between depression and anxiety scores as a function of their clinical variables. Our analyses reveal significant heterogeneity associated with gender and nervous-system medication use, which varies with age, revealing the complex dynamic relationship between depression and anxiety in bipolar disorders.

We study the complexity of smoothed agnostic learning, recently introduced by~\cite{CKKMS24}, in which the learner competes with the best classifier in a target class under slight Gaussian perturbations of the inputs. Specifically, we focus on the prototypical task of agnostically learning halfspaces under subgaussian distributions in the smoothed model. The best known upper bound for this problem relies on $L_1$-polynomial regression and has complexity $d^{\tilde{O}(1/\sigma^2) \log(1/\epsilon)}$, where $\sigma$ is the smoothing parameter and $\epsilon$ is the excess error. Our main result is a Statistical Query (SQ) lower bound providing formal evidence that this upper bound is close to best possible. In more detail, we show that (even for Gaussian marginals) any SQ algorithm for smoothed agnostic learning of halfspaces requires complexity $d^{\Omega(1/\sigma^{2}+\log(1/\epsilon))}$. This is the first non-trivial lower bound on the complexity of this task and nearly matches the known upper bound. Roughly speaking, we show that applying $L_1$-polynomial regression to a smoothed version of the function is essentially best possible. Our techniques involve finding a moment-matching hard distribution by way of linear programming duality. This dual program corresponds exactly to finding a low-degree approximating polynomial to the smoothed version of the target function (which turns out to be the same condition required for the $L_1$-polynomial regression to work). Our explicit SQ lower bound then comes from proving lower bounds on this approximation degree for the class of halfspaces.

We introduce cyclinbayes, an open-source R package for discovering linear causal relationships with both acyclic and cyclic structures. The package employs scalable Bayesian approaches with spike-and-slab priors to learn directed acyclic graphs (DAGs) and directed cyclic graphs (DCGs) under non-Gaussian noise. A central feature of cyclinbayes is comprehensive uncertainty quantification, including posterior edge inclusion probabilities, posterior probabilities of network motifs, and posterior probabilities over entire graph structures. Our implementation addresses two limitations in existing software: (1) while methods for linear non-Gaussian DAG learning are available in R and Python, they generally lack proper uncertainty quantification, and (2) reliable implementations for linear non-Gaussian DCG remain scarce. The package implements computationally efficient hybrid MCMC algorithms that scale to large datasets. Beyond uncertainty quantification, we propose a new decision-theoretic approach to summarize posterior samples of graphs, yielding principled point estimates based on posterior expected loss such as posterior expected structural Hamming distance and structural intervention distance. The package, a supplementary material, and a tutorial are available on GitHub at https://github.com/roblee01/cyclinbayes.

In safety-critical classification, the cost of failure is often asymmetric, yet Bayesian deep learning summarises epistemic uncertainty with a single scalar, mutual information (MI), that cannot distinguish whether a model's ignorance involves a benign or safety-critical class. We decompose MI into a per-class vector $C_k(x)=\sigma_k^{2}/(2\mu_k)$, with $\mu_k{=}\mathbb{E}[p_k]$ and $\sigma_k^2{=}\mathrm{Var}[p_k]$ across posterior samples. The decomposition follows from a second-order Taylor expansion of the entropy; the $1/\mu_k$ weighting corrects boundary suppression and makes $C_k$ comparable across rare and common classes. By construction $\sum_k C_k \approx \mathrm{MI}$, and a companion skewness diagnostic flags inputs where the approximation degrades. After characterising the axiomatic properties of $C_k$, we validate it on three tasks: (i) selective prediction for diabetic retinopathy, where critical-class $C_k$ reduces selective risk by 34.7% over MI and 56.2% over variance baselines; (ii) out-of-distribution detection on clinical and image benchmarks, where $\sum_k C_k$ achieves the highest AUROC and the per-class view exposes asymmetric shifts invisible to MI; and (iii) a controlled label-noise study in which $\sum_k C_k$ shows less sensitivity to injected aleatoric noise than MI under end-to-end Bayesian training, while both metrics degrade under transfer learning. Across all tasks, the quality of the posterior approximation shapes uncertainty at least as strongly as the choice of metric, suggesting that how uncertainty is propagated through the network matters as much as how it is measured.

Vector-quantized representations enable powerful discrete generative models but lack semantic structure in token space, limiting interpretable human control. We introduce SOM-VQ, a tokenization method that combines vector quantization with Self-Organizing Maps to learn discrete codebooks with explicit low-dimensional topology. Unlike standard VQ-VAE, SOM-VQ uses topology-aware updates that preserve neighborhood structure: nearby tokens on a learned grid correspond to semantically similar states, enabling direct geometric manipulation of the latent space. We demonstrate that SOM-VQ produces more learnable token sequences in the evaluated domains while providing an explicit navigable geometry in code space. Critically, the topological organization enables intuitive human-in-the-loop control: users can steer generation by manipulating distances in token space, achieving semantic alignment without frame-level constraints. We focus on human motion generation - a domain where kinematic structure, smooth temporal continuity, and interactive use cases (choreography, rehabilitation, HCI) make topology-aware control especially natural - demonstrating controlled divergence and convergence from reference sequences through simple grid-based sampling. SOM-VQ provides a general framework for interpretable discrete representations applicable to music, gesture, and other interactive generative domains.

High-dimensional datasets are frequently subject to contamination by outliers and heavy-tailed noise, which can severely bias standard regularized estimators like the Lasso. While Maximum Mean Discrepancy (MMD) has recently been introduced as a “universal” framework for robust regression, its application to high-dimensional Generalized Linear Models (GLMs) remains largely unexplored, particularly regarding variable selection. In this paper, we propose a penalized MMD framework for robust estimation and feature selection in GLMs. We introduce an $\ell_1$-penalized MMD objective and develop two versions of the estimator: a full $O(n^2)$ version and a computationally efficient $O(n)$ approximation. To solve the resulting non-convex optimization problem, we employ an algorithm based on the Alternating Direction Method of Multipliers (ADMM) combined with AdaGrad. Through extensive simulation studies involving Gaussian linear regression and binary logistic regression, we demonstrate that our proposed methods are highly competitive with classical penalized GLMs and existing robust benchmarks. Our approach shows particular resilience in maintaining a balance between estimation accuracy and variable selection across diverse contamination scenarios, especially in handling high-leverage points and heavy-tailed error distributions where traditional methods may fluctuate in performance.

This paper presents enhancements to the projection pursuit tree classifier and visual diagnostic methods for assessing their impact in high dimensions. The original algorithm uses linear combinations of variables in a tree structure where depth is constrained to be less than the number of classes – a limitation that proves too rigid for complex classification problems. Our extensions improve performance in multi-class settings with unequal variance-covariance structures and nonlinear class separations by allowing more splits and more flexible class groupings in the projection pursuit computation. Proposing algorithmic improvements is straightforward; demonstrating their actual utility is not. We therefore develop two visual diagnostic approaches to verify that the enhancements perform as intended. Using high-dimensional visualization techniques, we examine model fits on benchmark datasets to assess whether the algorithm behaves as theorized. An interactive web application enables users to explore the behavior of both the original and enhanced classifiers under controlled scenarios. The enhancements are implemented in the R package PPtreeExt.

Experimental evaluations of public policies often randomize a new intervention within many sites or blocks. After an overall statistically significant result is reported, the natural question from a policy maker is: where did effects occur? Standard adjustments for multiple testing answer this question with little power because they ignore how the experiment is organized: blocks nest within cohorts, sites, and districts. We organize the hypotheses in the shape of a tree that follows this administrative structure and test them top-down, stopping at any branch where the null is not rejected. A stopping rule and valid tests at each node suffice for weak control of the family-wise error rate (FWER). Whether the unadjusted procedure also controls the FWER in the strong sense depends on an error load computable from design quantities before any data are tested; when the load exceeds one, an adaptive $\alpha$-schedule, which we prove controls the FWER on regular and irregular trees without pruning, restores control. [Correction, August 2026: an anonymous referee identified errors in the previous version. The error loads reported in the paper were computed incorrectly; corrected loads exceed 1 in all 25 block-randomized MDRC education trials at the planning effect size $d = 0.20$, so the adjustment the earlier version claimed unnecessary is in fact required. The theorem claiming FWER control under branch pruning and its switching corollary are false as stated and are withdrawn; an exact counterexample attains FWER 0.063 at $\alpha = 0.05$. The detection comparison with the Hommel procedure is under recomputation. A correction notice on page 1 details what changes and what stands; a fully corrected version is in preparation.]

In many applications, weighted networks are constructed based on time series data: each time series is associated to a vertex and edge weights are given by pairwise correlations. The result is a network whose edge dependency structure violates the assumptions of most common network models. Nonetheless, it is common to analyze these "correlation networks" using embedding methods derived from edge-independent network models, based on a belief that the edges are approximately independent. In this work, we put this modeling choice on firm theoretical ground. We show that when the time series are expressible in terms of a small number of Fourier basis elements (or in some other suitably-chosen basis), correlation networks correspond to latent space networks with dependent edge noise in which the vertex-level latent variables encode the basis coefficients. Further, we show that when time series are observed subject to noise, spectral embedding of the resulting noisy correlation network still recovers these true vertex-level latent representations under suitable assumptions. This characterization of embeddings as learning Fourier coefficients appears to be folklore in the signal processing community in the context of principal component analysis, but is, to the best of our knowledge, new to the statistical network analysis literature.