Papers

Dirichlet process mixtures describe variation in effects or latent traits across sites, studies, or examinees. The concentration parameter controls the number and relative sizes of the clusters, but its hyperprior is often chosen by default. We describe a method for choosing this hyperprior when the number of units is fixed by design. Analysts specify an expected number of clusters and their uncertainty about that count. We translate these judgments into a hyperprior and examine what it implies about the sizes of the clusters. When the count and size judgments conflict, the Dual-Anchor procedure chooses and reports the trade-off. Simulations show that the default hyperprior can favor a single cluster when estimates for individual units are noisy. Applications to a multisite trial, a meta-analysis, and a vocabulary test show sensitivity in heterogeneity summaries despite relatively small changes in individual estimates. The DPprior R package implements the calibration and diagnostics.

Given that machine learning algorithms are increasingly being deployed to aid in high stakes decision-making, uncertainty quantification methods that wrap around these black box models such as conformal prediction have received much attention in recent years. In sequential settings, where data are observed/generated in a streaming fashion, traditional conformal methods do not provide any guarantee without fixing the sample size. More importantly, traditional conformal methods cannot cope with sequentially updated predictions. As such, we develop an extension of the conformal prediction and related probably approximately correct (PAC) prediction frameworks to sequential settings where the number of data points is not fixed in advance. The resulting prediction sets are anytime-valid in that their expected coverage is at the required level at any time chosen by the analyst even if this choice depends on the data. We present theoretical guarantees for our proposed methods and demonstrate their validity and utility on simulated and real datasets.

We address the problem of training conversion prediction models in advertising domains under privacy constraints, where direct links between ad clicks and conversions are unavailable. Motivated by privacy-preserving browser APIs and the deprecation of third-party cookies, we study a setting where the learner observes a sequence of clicks and a sequence of conversions, but can only link a conversion to a set of candidate clicks (an attribution set) rather than a unique source. We formalize this as learning from attribution sets generated by an oblivious adversary equipped with a prior distribution over the candidates. Despite the lack of explicit labels, we construct an unbiased estimator of the population loss from these coarse signals via a novel approach. Leveraging this estimator, we show that Empirical Risk Minimization achieves generalization guarantees that scale with the informativeness of the prior and is also robust against estimation errors in the prior, despite complex dependencies among attribution sets. Simple empirical evaluations on standard datasets suggest our unbiased approach significantly outperforms common industry heuristics, particularly in regimes where attribution sets are large or overlapping.

Starting from the probability distribution of finite N-body systems, which maximises the Havrda–Charvát entropy, we build a Stein-type goodness-of-fit test. The Maxwell–Boltzmann distribution is exact only in the thermodynamic limit, where the system is composed of infinitely many particles as N approaches infinity. For an isolated system with a finite number of particles, the equilibrium velocity distribution is compact and markedly non-Gaussian, being restricted by the fixed total energy. Using Stein's method, we first obtain a differential operator that characterises the target density. Its eigenfunctions are symmetric Jacobi polynomials, whose orthogonality yields a simple, parameter-free statistic. Under the null hypothesis that the data follows the finite-N distribution, the statistic converges to a chi-squared law, so critical values are available in closed form. Large-scale Monte Carlo experiments confirm exact size control and give a clear picture of the power. These findings quantify how quickly a finite system approaches the classical limit and provide a practical tool for testing kinetic models in regimes where normality cannot be assumed.

We study the problem of offline changepoint localization in a distribution-free setting. One observes a vector of data with a single changepoint, assuming that the data before and after the changepoint are i.i.d. (or more generally exchangeable) from arbitrary and unknown distributions. The goal is to produce a finite-sample confidence set for the index at which the change occurs without making any other assumptions. Existing methods often rely on parametric assumptions, tail conditions, or asymptotic approximations, or only produce point estimates. In contrast, our distribution-free algorithm, CONformal CHangepoint localization (CONCH), only leverages exchangeability arguments to construct confidence sets with finite sample coverage. By proving a conformal Neyman–Pearson lemma, we derive principled score functions that yield informative (small) sets. Moreover, with such score functions, the normalized length of the confidence set shrinks to zero under weak assumptions. We also establish a universality result showing that any distribution-free changepoint localization method must be an instance of CONCH. Experiments suggest that CONCH delivers precise confidence sets even in challenging settings involving images or text.

Evidence-informed policy on infections requires estimates of their effects on health. However, pathogenic variation, whereby occurrence of adverse outcomes depends on the infecting strain, might complicate the study of many infectious agents. Here, we consider the interpretation of epidemiologic studies on effects of infections on health when there is heterogeneity in strain-specific effects and information on strain composition is unavailable. We use potential outcomes and causal inference theory for analyses in the presence of multiple versions of treatment to argue that oft-reported quantities in these studies have a causal interpretation that depends on population frequencies of infecting strains. Moreover, as in other contexts where the treatment-variation-irrelevance assumption might be violated, transportability requires additional considerations, beyond those needed for non-compound exposures. This discussion, that considers potential heterogeneity in strain-specific effects, will facilitate interpretation of these studies, and for the reasons mentioned above, also highlights the value of pathogen subtype data.

Neural-network techniques are often transferred across architecture families by analogy, but such transfer is valid only when the assumptions required by a technique are preserved. We introduce this idea as inheritance between model classes. Using a unified node-level framework with tensor-valued activations, we prove that generalized feedforward networks (GFFNs) form a strict subset of generalized convolutional networks (GCNNs), so GCNN properties transfer directly to GFFNs. The reverse direction is not automatic: standard CNN nodes use spatial kernels, while FFN nodes use one scalar weight per input contribution. We introduce model projection to recover a restricted reverse inheritance path. Projection freezes each convolutional input-channel sub-function and learns one scalar coefficient for each input-output channel contribution, giving projected CNN nodes the GFFN-style trainable structure of scalar-weighted input recombination. This inherited structure leads naturally to parameter-efficient transfer learning. Across multiple ImageNet-pretrained CNN backbones and downstream image-classification datasets, model projection is competitive with standard and PEFT baselines and provides an effective initialization for subsequent full fine-tuning.

Precision medicine seeks to match patients with treatments that produce the greatest benefit. The Predicted Individual Treatment Effect (PITE)-the difference between predicted outcomes under treatment and control-quantifies this benefit but is difficult to estimate due to unobserved counterfactuals, high dimensionality, and complex interactions. We compared 30+ modeling strategies, including penalized and projection-based methods, flexible learners, and tree-ensembles, using a structured simulation framework varying sample size, dimensionality, multicollinearity, and interaction complexity. Performance was measured using root mean squared error (RMSE) for prediction accuracy and directional accuracy (DIR) for correctly classifying benefit versus harm. Internal validation produced optimistic estimates, whereas external validation with distributional shifts and higher-order interactions more clearly revealed model weaknesses. Penalized and projection-based approaches-ridge, lasso, elastic net, partial least squares (PLS), and principal components regression (PCR)-consistently achieved strong RMSE and DIR performance. Flexible learners excelled only under strong signals and sufficient sample sizes. Results highlight robust linear/projection defaults and the necessity of rigorous external validation.

Expand-and-sparsify representations are a class of theoretical models that capture sparse representation phenomena observed in the sensory systems of many animals. At a high level, these representations map an input $x \in \mathbb{R}^d$ to a much higher dimension $m \gg d$ via random linear projections before zeroing out all but the $k \ll m$ largest entries. The result is a $k$-sparse vector in $\{0,1\}^m$. We study the suitability of this representation for two fundamental statistical problems: density estimation and mode estimation. For density estimation, we show that a simple linear function of the expand-and-sparsify representation produces an estimator with minimax-optimal $\ell_{\infty}$ convergence rates. In mode estimation, we provide simple algorithms on top of our density estimator that recover single or multiple modes at optimal rates up to logarithmic factors under mild conditions.

Diffusion models generate samples through a sequence of learned denoising steps, and recent work has studied how semantic structure appears along this sampling process. We study this question in deterministic samplers by measuring semantic accessibility: how much information about a final semantic property, such as an image class label or attribute, can be extracted from the seed and intermediate states along the trajectory that produces the sample. Using DDIM sampling, for which each initial noise seed determines a unique trajectory and final image, we train separate classifiers (probes) at several points along the trajectory to predict a semantic property of the final image. We measure how well such a property can be predicted from the state at a given point using top-1 accuracy and normalized mutual information. Across MNIST, Fashion-MNIST, CIFAR-10, and CelebA, class labels and image attributes can be predicted above chance from the initial noise seed, and along DDIM trajectories, this accessibility exceeds matched-noise forward baselines. We find that semantic accessibility is substantially higher along the trajectories whose final images are classified with high confidence than along those classified with low confidence. These measurements provide a quantitative view of when semantic properties can be recovered during deterministic generation.

Logistic regression is the most commonly used method for constructing predictive models for binary responses. One significant drawback to this approach, however, is that the asymptotes of the logistic response function are fixed at 0 and 1, and there are many applications for which this constraint is inappropriate. More flexible models have been proposed for this application, most proceeding by supplementing the logistic response function with additional parameters. In this article we extend these models to allow correlated responses and the inclusion of covariates. This is achieved through the compound logistic regression model, for which the mean response is a function of several logistic regression functions. This permits a greater variety of models, while retaining the advantages of logistic regression.

Effective population size (Ne(t)) is a fundamental parameter in population genetics and phylodynamics that quantifies genetic diversity and reveals demographic history. Coalescent-based methods enable the inference of Ne(t) trajectories through time from phylogenies reconstructed from molecular sequence data. Understanding the ecological and environmental drivers of population dynamics requires linking Ne(t) to external covariates. Existing approaches typically impose log-linear relationships between covariates and Ne(t), which may fail to capture complex biological processes and can introduce bias when the true relationship is nonlinear. We present a flexible Bayesian framework that integrates covariates into coalescent models with piecewise-constant Ne(t) through a Gaussian process (GP) prior. The GP, a distribution over functions, naturally accommodates nonlinear covariate effects without restrictive parametric assumptions. This formulation improves estimation of covariate-Ne(t) relationships, mitigates bias under nonlinear associations, and yields interpretable uncertainty quantification that varies across the covariate space. To balance global covariate-driven patterns with local temporal dynamics, we couple the GP prior with a Gaussian Markov random field that enforces smoothness in Ne(t) trajectories. Through simulation studies and three empirical applications - yellow fever virus dynamics in Brazil (2016-2018), late-Quaternary musk ox demography, and HIV-1 CRF02-AG evolution in Cameroon - we demonstrate that our method both confirms linear relationships where appropriate and reveals nonlinear covariate effects that would otherwise be missed or mischaracterized. This framework advances phylodynamic inference by enabling more accurate and biologically realistic modeling of how environmental and epidemiological factors shape population size through time.

In recent work it is shown that Q-learning with linear function approximation is stable, in the sense of bounded parameter estimates, under the $(\varepsilon,\kappa)$-tamed Gibbs policy; $\kappa$ is inverse temperature, and $\varepsilon>0$ is introduced for additional exploration. Under these assumptions it also follows that there is a solution to the projected Bellman equation (PBE). Left open is uniqueness of the solution, and criteria for convergence outside of the standard tabular or linear MDP settings. The present work extends these results to other variants of Q-learning, and clarifies prior work: a one dimensional example shows that under an oblivious policy for training there may be no solution to the PBE, or multiple solutions, and in each case the algorithm is not stable under oblivious training. The main contribution is that far more structure is required for convergence. An example is presented for which the basis is ideal, in the sense that the true Q-function is in the span of the basis. However, there are two solutions to the PBE under the greedy policy, and hence also for the $(\varepsilon,\kappa)$-tamed Gibbs policy for all sufficiently small $\varepsilon>0$ and $\kappa\ge 1$.

Reliable preparation of many-body ground states is an essential task in quantum computing, with applications spanning areas from chemistry and materials modeling to quantum optimization and benchmarking. A variety of approaches have been proposed to tackle this problem, including variational methods. However, variational training often struggle to navigate complex energy landscapes, frequently encountering suboptimal local minima or suffering from barren plateaus. In this work, we introduce an iterative strategy for ground-state preparation based on a stepwise (discretized) Hamiltonian deformation. By complementing the Variational Quantum Eigensolver (VQE) with adiabatic principles, we demonstrate that solving a sequence of intermediate problems facilitates tracking the ground-state manifold toward the target system, even as we scale the system size. We provide a rigorous theoretical foundation for this approach, proving a lower bound on the loss variance that suggests trainability throughout the deformation, provided the system remains away from gap closings. Numerical simulations, including the effects of shot noise, confirm that this path-dependent tracking consistently converges to the target ground state.

Mpox is an orthopoxvirus that infects humans and animals and is transmitted primarily through close physical contact. The episodic and spatially heterogeneous dynamics of Mpox transmission underscores the need for timely, area-specific forecasts to support targeted public health responses in the U.S. We develop a Vector Autoregression model with Lasso regularization (VAR-Lasso) to generate rolling two-week-ahead forecasts of weekly Mpox cases for eight high-incidence U.S. jurisdictions using national surveillance data from the Centers for Disease Control and Prevention (CDC). The VAR-Lasso model identifies significant long-lag, cross-jurisdictional predictors. For a case study in San Diego County (SDC), these statistical predictors align with phylogenetic analysis that traces a 2023 cluster in SDC to an outbreak in Illinois six months earlier. As the need for public health action is often greatest when incidence is increasing, our performance evaluation focuses on positive-slope weighted error metrics. Forecast performance of the VAR-Lasso model is compared to a uni-variate Auto-Regressive (AR) Lasso model and a naive moving-average estimate. The models are compared using slope-weighted Root Mean Squared Error (RMSE), slope-weighted Mean Absolute Error (MAE), and slope-weighted bias. Across all observations, the VAR-Lasso model reduces slope-weighted RMSE, MAE, and bias by 12%, 7%, and 66% relative to the AR model, and by 16%, 13%, and 76% relative to the naive benchmark. Our findings highlight the value of sparse multivariate time-series models that leverage cross-jurisdictional case data for early forecasting of Mpox outbreaks. Such forecasting can aid health departments in proactively providing timely resources and messaging to mitigate the risks of a future outbreak.

We study robustness verification of neural networks via metric algebraic geometry. For polynomial neural networks, certifying a robustness radius amounts to computing the distance to the algebraic decision boundary. We use the Euclidean distance (ED) degree as an intrinsic measure of the complexity of this problem, analyze the associated ED discriminant, and introduce a parameter discriminant that detects parameter values at which the ED degree drops. We derive formulas for the ED degree for several network architectures and characterize the expected number of real critical points in the infinite-width limit. We develop symbolic elimination methods to compute these quantities and homotopy-continuation methods for exact robustness certification. Finally, experiments on lightning self-attention modules reveal decision boundaries with strictly smaller ED degree than generic cubic hypersurfaces of the same ambient dimension.

Many engineering and scientific workflows rely on expensive black-box evaluations, requiring sequential decisions that must both improve task performance and reduce uncertainty. Bayesian optimization (BO) and Bayesian experimental design (BED) provide powerful but largely separate treatments of goal-directed optimization and information-seeking experimentation, leaving limited guidance for hybrid settings in which learning and optimization are intrinsically coupled. We propose Pragmatic Curiosity (PraC), a unified framework for hybrid learning and optimization via active inference. PraC evaluates candidate queries by trading information gain about a task-relevant latent symbol against an expected regret-based potential over outcomes. This formulation exposes three operational design choices: which latent quantity should be clarified, how task value is encoded as regret, and how strongly information gain should be exchanged against pragmatic value. We instantiate PraC across three regimes of increasing complexity: decision-oriented plume monitoring with fixed global symbols and known downstream losses, targeted active search with induced local symbols and evolving coverage goals, and composite Bayesian optimization with hierarchical regret learning under unknown preferences. Across these regimes, PraC reduces downstream decision risk, improves coverage of critical outcome regions, and jointly learns predictive and preference structures without relying on task-specific staging rules.

We study a data-dependent notion of diffusion-model generalization: when a model does not memorize the training set, where do its generated samples go relative to the geometry induced by the data? To answer this, we introduce a time-dependent family of log-density ridge manifolds constructed from the smoothed empirical distribution, and use it to characterize reverse-time inference. Our main result shows that generated samples evolve by a reach-align-slide mechanism: they first enter a neighborhood of the ridge, then their distance to the ridge is controlled by the normal component of training error, and finally their motion along the ridge is controlled by the tangential component. We further connect this geometric picture to training dynamics through directional decompositions of the learned error, and make this link explicit for random feature models, where architectural bias and optimization error can be separated quantitatively. Experiments on synthetic multimodal data and MNIST latent diffusion support the predicted geometric behavior in both low and high dimensions.

Thompson sampling (TS) is widely used for stochastic multi-armed bandits, yet its inferential properties under adaptive data collection are subtle. Classical asymptotic theory for sample means can fail because arm-specific sample sizes are random and coupled with the rewards through the action-selection rule. We study adaptive inference for Thompson sampling with Gaussian randomized indices in $K$-armed stochastic bandits with independent sub-Gaussian reward noises, and identify optimism as a key mechanism for restoring stability, meaning that each arm's pull count concentrates around a deterministic scale. This stability yields asymptotically valid Wald inference despite adaptive sampling. First, we prove that variance-inflated TS is stable for any $K \ge 2$, including the challenging regime where multiple arms are optimal, with asymptotically uniform allocation over optimal arms and sharp logarithmic pull-count asymptotics for suboptimal arms. This resolves the $K$-armed extension question raised by \citet{halder2025stable}, using new winner-map and Lyapunov-drift techniques to control allocation among multiple optimal arms. Second, we analyze an alternative optimistic modification that keeps the Gaussian index variance unchanged but adds an explicit mean bonus to the index center, and establish a similar stability conclusion. In summary, suitably implemented optimism stabilizes Thompson sampling and enables asymptotically valid Wald inference in multi-armed bandits, while incurring only a mild additional regret cost.

We consider a causal inference problem frequently encountered in online advertising systems, where a publisher (e.g., Instagram, TikTok) interacts repeatedly with human users and advertisers by sporadically displaying to each user an advertisement selected through an auction. Each treatment corresponds to a parameter value of the advertising mechanism (e.g., auction reserve-price), and we want to estimate through experiments the corresponding long-term treatment effect (e.g., annual advertising revenue). In our setting, the treatment affects not only the instantaneous revenue from showing an ad, but also changes each user's interaction-trajectory, and each advertiser's bidding policy – as the latter is constrained by a finite budget. In particular, each a treatment may even affect the size of the population, since users interact longer with a tolerable advertising mechanism. We drop the classical i.i.d. assumption and model the experiment measurements (e.g., advertising revenue) as a stopped random walk, and use a budget-splitting experimental design, the Anscombe Theorem, a Wald-like equation, and a Central Limit Theorem to construct confidence intervals for the long-term treatment effect.

Softmax Self-Attention (SSA) is a key component of Transformer architectures. However, when utilised within skipless architectures, which aim to improve representation learning, recent work has highlighted the inherent instability of SSA due to inducing rank collapse and poorly-conditioned Jacobians. In this work, we design a novel attention mechanism: Orthogonal Self-Attention (OSA), which aims to bypass these issues with SSA, in order to allow for (non-causal) Transformers without skip connections and normalisation layers to be more easily trained. In particular, OSA parametrises the attention matrix to be orthogonal via mapping a skew-symmetric matrix, formed from query-key values, through the matrix exponential. We show that this can be practically implemented, by exploiting the low-rank structure of our query-key values, resulting in the computational complexity and memory cost of OSA scaling linearly with sequence length. Furthermore, we derive an initialisation scheme for which we prove ensures that the Jacobian of OSA is well-conditioned.

Neural scaling laws relate loss to model size in large language models (LLMs), yet depth and width may contribute to performance differently, requiring more detailed studies. Here, we quantify how depth affects loss via analysis of LLMs and toy residual networks. We find loss scales inversely proportional to depth in LLMs, probably due to functionally similar layers reducing error through ensemble averaging rather than compositional learning or discretizing smooth dynamics. This regime is inefficient yet robust and may arise from the architectural bias of residual networks and target functions incompatible with smooth dynamics. The findings suggest that improving LLM efficiency may require architectural innovations to encourage compositional use of depth.

Sampling from a distribution $p(x) \propto e^{-\mathcal{E}(x)}$ known up to a normalising constant is an important and challenging problem in statistics. Recent years have seen the rise of a new family of amortised sampling algorithms, commonly referred to as diffusion samplers, that enable fast and efficient sampling from an unnormalised density. Such algorithms have been widely studied for continuous-space sampling tasks; however, their application to problems in discrete space remains largely unexplored. Although some progress has been made in this area, discrete diffusion samplers do not take full advantage of ideas commonly used for continuous-space sampling. In this paper, we propose to bridge this gap by introducing off-policy training techniques for discrete diffusion samplers. We show that these techniques improve the performance of discrete samplers on both established and new synthetic benchmarks. Next, we generalise discrete diffusion samplers to the task of bridging between two arbitrary distributions, introducing data-to-energy Schrödinger bridge training for the discrete domain for the first time. Lastly, we showcase the application of the proposed diffusion samplers to data-free posterior sampling in the discrete latent spaces of image generative models.

Recent work shows that preference alignment objectives can be interpreted as divergence estimators between aligned (preferred) & unaligned (less-preferred) distributions, yielding a principled recipe for designing alignment losses. However, this view has so far been limited to preference-based supervision. We extend it to general LLM alignment, including reinforcement learning with verifiable rewards (RLVR), where alignment feedback is given only as scalar rewards. We introduce $f$-Group Relative Policy Optimization ($f$-GRPO), a class of on-policy RL objectives, and $f$-Hybrid Alignment Loss ($f$-HAL), which combines on-policy reward optimization with off-policy preference supervision. We show that these objectives estimate $f$-divergences between reward-aligned & reward-unaligned distributions induced by above- & below-average reward responses, and prove expected reward improvement after alignment. Empirically, $f$-GRPO improves over GRPO on math-reasoning RLVR tasks, while hybrid $f$-HAL mitigates reward hacking in on-policy safety alignment when verifiable rewards are unavailable and learned reward models must be used.

Recent evidence suggests that analyzing the presence/absence of taxonomic features can offer a compelling alternative to differential abundance analysis in microbiome studies. However, standard approaches to differential prevalence analysis face challenges with boundary cases and multiple testing. To address these limitations, we developed DiPPER (Differential Prevalence via Probabilistic Estimation in R), a method based on Bayesian hierarchical modeling. We benchmarked our method against existing differential prevalence methods, along with two differential abundance tools, using publicly available data from 57 human gut microbiome studies. We observed considerable variation in performance across the evaluated methods. Importantly, DiPPER demonstrated high sensitivity to detect potentially differentially prevalent features while maintaining a well-calibrated family-wise error rate under the global null hypothesis. Most notably, it outperformed the alternatives in the replication of findings across independent studies. Furthermore, DiPPER provides differential prevalence estimates and uncertainty intervals that are inherently adjusted for multiple testing.

Transformers underpin modern large language models (LLMs) and are commonly assumed to be behaviorally unstructured at random initialization, with all meaningful preferences emerging only through large-scale training. We challenge this assumption by showing that randomly initialized transformers already exhibit strong and systematic structural biases. In particular, untrained models display extreme token preferences: across random input sequences, certain tokens are predicted with probabilities orders of magnitude larger. We provide a mechanistic explanation for this phenomenon by dissecting the transformer architecture at initialization. We show that extreme token preference arises from a contraction of token representations along a random seed-dependent direction. This contraction is driven by two interacting forces: (i) asymmetric nonlinear activations in MLP sublayers induce global (inter-sequence) representation concentration, and (ii) self-attention further amplifies this effect through local (intra-sequence) aggregation. Together, these mechanisms align hidden representations along a direction determined solely by the random initialization, producing highly non-uniform next-token predictions. Beyond mechanistic insight, we demonstrate that these initialization-induced biases persist throughout training, forming a stable and intrinsic model identity. Leveraging this property, we introduce SeedPrint, a fingerprinting method that can reliably distinguish models that differ only in their random initialization, even after extensive training and under substantial distribution shift. Finally, we identify a fundamental positional discrepancy inherent to the attention mechanism's intra-sequence contraction that is causally linked to the attention-sink phenomenon. This discovery provides a principled explanation for the emergence of sinks and offers a pathway for their control.

We introduce Wedge Sampling, a new non-adaptive sampling scheme for low-rank tensor completion. We study recovery of an order-$k$ low-rank tensor of dimension $n\times\cdots\times n$ from structured observations of its entries. Unlike the standard uniform entry model (i.e., i.i.d. samples from $[n]^k$), wedge sampling allocates observations to structured length-two patterns (wedges) in an associated bipartite sampling graph. By directly promoting these length-two connections, the sampling design strengthens the spectral signal that underlies efficient initialization, in regimes where uniform sampling is too sparse to generate enough informative correlations. Our main result shows that this change in sampling paradigm enables polynomial-time algorithms to achieve both weak and exact recovery with nearly linear sample complexity in $n$. The approach is also plug-and-play: wedge-sampling-based spectral initialization can be combined with existing refinement procedures (e.g., spectral or gradient-based methods) using only an additional $\tilde O(n)$ uniformly sampled entries, substantially improving over the $\tilde O(n^{k/2})$ sample complexity typically required under uniform entry sampling for efficient methods. We also formulate a noisy wedge-sampling extension for additive Gaussian observations and analyze both the spectral and gradient-descent procedures under suitable signal-to-noise conditions. Thus, the computational barrier in tensor completion is sensitive to the observation model: while it persists under uniform entry sampling, it can be bypassed by non-adaptive structured designs that provide a stronger initialization.

Two-sample testing, where we aim to determine whether two distributions are equal or not equal based on samples from each one, is challenging if we cannot place assumptions on the properties of the two distributions. In particular, certifying equality of distributions, or even providing a tight upper bound on the total variation (TV) distance between the distributions, is impossible to achieve in a distribution-free regime. In this work, we examine the blurred TV distance, a relaxation of TV distance that enables us to perform inference without assumptions on the distributions. We provide theoretical guarantees for distribution-free upper and lower bounds on the blurred TV distance, and examine its properties in high dimensions.

Graph-structured data is ubiquitous and powerful in representing complex relationships in many online platforms. While graph neural networks (GNNs) are widely used to learn from such data, counterfactual graph learning has emerged as a promising approach to improve model interpretability. Counterfactual explanation research focuses on identifying a counterfactual graph that is similar to the original but leads to different predictions. These explanations optimize two objectives simultaneously: the sparsity of changes in the counterfactual graph and the validity of its predictions. Building on these qualitative optimization goals, this paper introduces CFRecs, a novel framework that transforms counterfactual explanations into actionable insights. CFRecs employs a two-stage architecture consisting of a graph neural network (GNN) and a graph variational auto-encoder (Graph-VAE) to strategically propose minimal yet high-impact changes in graph structure and node attributes to drive desirable outcomes in recommender systems. We apply CFRecs to Zillow's graph-structured data to deliver actionable recommendations for both home buyers and sellers with the goal of helping them navigate the competitive housing market and achieve their homeownership goals. Experimental results on Zillow's user-listing interaction data demonstrate the effectiveness of CFRecs, which also provides a fresh perspective on recommendations using counterfactual reasoning in graphs.

Community detection in networks is a fundamental problem in machine learning and statistical inference, with applications in social networks, biological systems, and communication networks. The stochastic block model (SBM) serves as a canonical framework for studying community structure, and exact recovery, identifying the true communities with high probability, is a central theoretical question. While classical results characterize the phase transition for exact recovery based solely on graph connectivity, many real-world networks contain additional data, such as node attributes or labels. In this work, we study exact recovery in the Data Block Model (DBM), an SBM augmented with node-associated data, as formalized by Asadi, Abbe, and Verdú (2017). We introduce the Chernoff–TV divergence and use it to characterize a sharp exact recovery threshold for the DBM. We further provide an efficient algorithm that achieves this threshold, along with a matching converse result showing impossibility below the threshold. Finally, simulations validate our findings and demonstrate the benefits of incorporating vertex data as side information in community detection.