arXiv subjects

Nadja Klein

Publications and source records attributed to Nadja Klein.

At least 19 recordsLinked to original sources

arXiv2026

XAI-Arena: Can LLMs Assess the Quality of XAI Explanations?

Evaluating the quality of explanations produced by explainable AI (XAI) methods remains challenging because existing approaches often rely on subjective human judgment, limiting reproducibility, scalability, and comparability between studies. We examine whether LLMs can serve as a reproducible and scalable mechanism to make comparative assessments of the quality of XAI explanations. We introduce XAI-Arena, an LLM-as-a-judge framework for scalable, reproducible, multidimensional, and stakeholder-sensitive evaluation of XAI explanation quality. XAI-Arena then allows us to compare XAI explanations along various dimensions, namely, perceived simplicity, clarity, task adequacy, trust calibration, actionability, transparency, faithfulness, and overall interpretability. We then benchmark XAI explanation methods across various datasets, machine learning models, and stakeholder personas. Human validation shows a strong positive association between LLM-generated and human ratings (Spearman's rho=.693, p<.001). Together, LLM-based evaluations can capture systematic differences in XAI explanation quality and provide a scalable and reproducible framework for comparative assessment of XAI explanations.

cs.AI↗

arXiv2026

Interpreting Protein Language Model Embeddings via Orthogonal Projection for Protein Fitness Prediction

Recently, there has been a growing adoption of protein language models (PLMs) in biomedical science. Their embeddings provide a rich numerical representation of protein sequences which achieve state-of-the-art performance on several downstream tasks including protein fitness prediction. However, PLM embeddings are not directly interpretable and, thereby, it remains unclear what features they encode. To gain insight into which biochemical properties of the protein are driving the prediction, we leverage an orthogonal projection technique that removes linear effects of known tabular features from embeddings and extend it to high-order and interaction effects. In this way, we remove the effects of interpretable biochemical features from PLM embeddings. In an ablation study, we show that this leads to a decrease in performance for a downstream classifier trained only on the embeddings to predict protein fitness. In an additional evaluation, we find that these biochemical features explain a substantial part of the variance in the predictions of this classifier. Hence, we can show that PLM embeddings encode patterns correlated with biochemical properties and quantify their contribution to predicting protein fitness. This computationally efficient approach is not limited to the features or embeddings considered here and is readily transferable to problem settings beyond protein fitness prediction.

cs.LG↗

arXiv2026

Active Interaction-Aware Model Predictive Path Integral via Ego-Conditioned Generative Predictions

Dense traffic is inherently interactive. The ego vehicle and surrounding agents continuously influence each other's reactions, making "what-if" reasoning essential for safe and efficient driving. To enable such an active interaction-aware behavior, we propose a planning framework that integrates an ego-conditioned generative autoregressive prediction model within Model Predictive Path Integral (MPPI) control. The generative prediction model outputs stochastic, multi-modal predictions of surrounding agents conditioned on each of the ego's considered future actions. A nested sampling scheme enables tractable evaluation of expected cost and collision risk under the induced distribution. This formulation allows the ego to actively probe how different candidate actions shape the interaction outcomes and to identify actions that reduce ambiguity in uncertain interactions. Closed-loop simulations demonstrate improved safety and efficiency compared to conventional predict-then-plan and passive interaction-aware approaches.

cs.RO↗

arXiv2026

Copulas for Geostatistical Data: Foundations, Modeling Principles and Statistical Inference

Spatial statistics commonly describes spatial dependence through second-order quantities such as covariance functions and variograms, often within Gaussian random-field models and under structural assumptions such as stationarity, isotropy, or distance-based decay. Copulas offer a complementary framework that separates marginal distributions from dependence and permits a broad range of non-Gaussian dependence structures. Because the finite-dimensional distributions of a spatial random field can always be decomposed into margins and copulas through Sklar's theorem, copulas provide a natural language for studying spatial dependence beyond second-order summaries. Yet the relevant literature has developed along several largely separate strands across spatial statistics, copula modeling, stochastic processes, and application domains, often with different terminology and modeling objectives. This review brings these strands together: We revisit classical concepts from spatial statistics through a copula lens, discuss copula-based tools for describing spatial dependence, and systematically review constructions of spatial copula models. Particular emphasis is placed on Kolmogorov consistency and on the distinction between models defined for a fixed set of locations and genuinely process-level constructions. We also discuss statistical inference, extensions to spatio-temporal settings, and emerging directions involving flexible marginal and dependence models. By clarifying the relationships among existing approaches and their respective strengths and limitations, the review provides a unified perspective on the interface between copula modeling and spatial statistics.

stat.ME↗

arXiv2026

Bayesian Effect Selection for Additive Quantile Regression with an Application to Air Pollution Thresholds

Air pollution regulatory limits are typically defined in terms of exceedances of concentration thresholds which are naturally related to conditional quantiles of the pollutant distribution and are therefore of direct relevance for assessing severe pollution events. At the same time, it is important to determine not only whether a covariate affects air pollution but also whether this effect is linear, nonlinear, or both. We address these issues by developing a Bayesian effect selection approach for additive quantile regression. While commonly used mixed model representations (MMRs) of penalized splines allow for flexible nonlinear effects, they do not provide a meaningful separation of linear and nonlinear effect components. We therefore employ a Demmler-Reinsch basis expansion, which yields an orthogonal decomposition of each additive effect into linear and nonlinear parts and show theoretically that both effect components can be estimated consistently. To facilitate data-driven model building, we propose Bayesian effect selection with separate spike and slab priors on the scalar importance parameters associated with the linear and nonlinear components and implement an efficient Gibbs sampler. Through simulation studies, we demonstrate robustness to the misspecification induced by the employed asymmetric Laplace working likelihood and show superior performance relative to the MMR. In a detailed analysis of air pollution data in Madrid, Spain we highlight the added value of flexibly modeling extreme nitrogen dioxide (NO$_2$) concentrations and reveal that threshold-relevant pollution levels are driven differently by climatological variables and traffic-related spatial structure. These findings underline the need for advanced statistical models that support short-term decision-making and help local authorities mitigate, or potentially prevent, exceedances of NO$_2$ concentration limits.

stat.ME↗

arXiv2026

Bayesian meta-learning for modeling Alzheimer's disease progression

Predicting whether an individual with Alzheimer's disease will experience mild or severe disease progression is essential for personalized treatment. Typically, practitioners seek to predict the distribution of a discrete disease score, conditional on an individual's current MRI volume and their historical disease trajectory. Classical statistical regression models and single-task neural networks are not well-suited for this purpose because fitting separate models is infeasible (since each individual typically has few observations), while ignoring individual-level correlation leads to poor generalization. Meta-learning, in contrast, provides a natural avenue to dynamically predict distributions without retraining and model nonlinear relationships between the outcome and covariates. Motivated by this, we propose a Bayesian meta-learner that is trained on multiple individuals but tailors the predictive disease score distribution to each individual's historical data. Our model predicts on unseen individuals without retraining, scales linearly with the number of historical observations, and is guaranteed to be less overconfident when predicting long-term disease scores compared to its deterministic counterpart. On real-world data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database, our model achieves performance competitive with both single-task models and deterministic meta-learners, while substantially improving performance when predicting long-term disease progression.

stat.ML↗

arXiv2026

Diffusion-based Denoising Beats Vanilla Score Matching in Parameter Estimation: A Theoretical Explanation

Score matching is an alternative to maximum likelihood estimation when the normalizing constant is unknown or too costly to evaluate. However, vanilla score matching has shown to be inefficient relative to maximum likelihood estimation for multimodal distributions with well-separated modes, which are commonly encountered in practical applications. We compare a novel diffusion-based denoising score matching estimator (DDSME) to the vanilla score matching estimator (SME) in this scenario. In particular, we prove statistical guarantees for both estimators, showing that the error bound for the vanilla SME worsens when the separation between the modes increases, which can be avoided in case of the DDSME with suitable hyperparameter tuning. This provides a novel theoretical explanation for the superior behavior of diffusion-based score matching over the vanilla version. We support our theoretical findings by numerical experiments.

stat.ML↗

arXiv2026

Point Cloud Sequence Encoding for Material-conditioned Graph Network Simulators

Graph Network Simulators (GNSs) have emerged as powerful surrogates for complex physics-based simulation, offering inherent differentiability and orders-of-magnitude speedups over traditional solvers. However, GNSs typically assume access to the underlying material parameters, such as stiffness or viscosity, severely limiting their utility in realistic experimental settings. While recent meta-learning approaches address the parameter dependency by inferring properties from mesh trajectories, reconstructing a mesh from an observed scene is challenging. In this work, we introduce Point Cloud Encoding for Accurate Context Handling (PEACH), a novel framework that applies in-context learning on point clouds to adapt a learned simulator to unseen physical properties during inference. Our approach relies on a novel spatio-temporal point cloud sequence encoder, as well as two forms of auxiliary supervision to help improve simulation fidelity. We demonstrate that PEACH is capable of accurate zero-shot sim-to-real transfer on a challenging, dynamic scene. Experiments on simulation scenes show that PEACH even outperforms mesh-based baselines on prediction accuracy, while being much more practical for real-world deployment.

cs.LG↗

arXiv2026

QDSB: Quantized Diffusion Schr\"odinger Bridges

Learning generative models in settings where the source and target distributions are only specified through unpaired samples is gaining in importance. Here, one frequently-used model are Schr\"odinger bridges (SB), which represent the most likely evolution between both endpoint distributions. To accelerate training, simulation-free SBs avoid the path simulation of the original SB models. However, learning simulation-free SBs requires paired data; a coupling of the source and target samples is obtained as the solution of the entropic optimal transport (OT) problem. As obtaining the optimal global coupling is infeasible in many practical cases, the entropic OT problem is iteratively solved on minibatches instead. Still, the repeated cost remains substantial and the locality can distort the global transport geometry. We propose quantized diffusion Schr\"odinger bridges (QDSB), which compute the endpoint coupling on anchor-quantized endpoint distributions and lift the resulting plan back to original data points through cell-wise sampling. We show that the regularized optimal coupling is stable w.r.t. anchor quantization, with an error controlled by the quality of the anchor approximation. In real-world experiments, QDSB matches the sample quality of existing baselines, requiring substantially less time. Code and data are available at github.com/mathefuchs/qdsb.

cs.LG↗

arXiv2026

From Model Uncertainty to Human Attention: Localization-Aware Visual Cues for Scalable Annotation Review

High-quality labeled data is essential for training robust machine learning models, yet obtaining annotations at scale remains expensive. AI-assisted annotation has therefore become standard in large-scale labeling workflows. However, in tasks where model predictions carry two independent components, a class label and spatial boundaries, a model may classify an object with high confidence while mislocalizing it. Existing AI-assisted workflows offer annotators no signal about where spatial errors are most likely. Without such guidance, humans may systematically underinspect subtly misplaced boxes. We address this by studying the effect of visualizing spatial uncertainty via a purpose-built interface. In a controlled study with 120 participants, those receiving uncertainty cues achieve higher label quality while being faster overall. A box-level analysis confirms that the cues redirect annotator effort toward high-uncertainty predictions and away from well-localized boxes. These findings establish localization uncertainty as a lever to improve human-in-the-loop annotation. Code is available at https://mos-ks.github.io/MUHA/.

cs.HC↗

arXiv2026

Probabilistic Object Detection with Conformal Prediction

Conformal Prediction (CP) is a distribution-free method for constructing prediction sets with marginal finite-sample coverage guarantees, making it a suitable framework for reliable uncertainty quantification in safety-critical object detection. However, object detection introduces structured multi-output predictions, complicating the application of classical CP theory developed for single outputs. In addition, standard, unscaled CP produces fixed-width prediction intervals across inputs, leading to unnecessary width for low-uncertainty predictions. While scaled CP addresses this by adapting the interval width to an input-dependent uncertainty estimate, prior work has neither systematically compared unscaled and scaled CP for multi-class object detection, nor integrated CP with a complementary uncertainty quantification method in this setting. We fill this gap by: (i) applying CP coordinate-wise to bounding box corners with a Bonferroni correction for box-level guarantees; (ii) scaling the resulting intervals using per-prediction aleatoric uncertainty estimates derived from a probabilistic object detector trained with loss attenuation, evaluated in uncalibrated and two calibrated variants; (iii) extending to a two-step pipeline that constructs prediction sets for the class using RAPS and conditions the conformalized bounding boxes on the predicted class set. Across three autonomous driving datasets (KITTI, BDD, CODA), including a cross-domain setting under distribution shift, scaled CP consistently improves interval sharpness over unscaled CP, achieving up to 19% higher IoU and 39% lower interval scores, without sacrificing coverage. Class-wise calibration further improves coverage for both variants with a negligible effect on sharpness. Together, these improvements yield more actionable uncertainty estimates for real-time, real-world object detection.

cs.CV↗

arXiv2026

Semi-structured multi-state delinquency model for mortgage default

We propose a semi-structured discrete-time multi-state model to analyse mortgage delinquency transitions. This model combines an easy-to-understand structured additive predictor, which includes linear effects and smooth functions of time and covariates, with a flexible neural network component that captures complex nonlinearities and higher-order interactions. To ensure identifiability when covariates are present in both components, we orthogonalise the unstructured part relative to the structured design. For discrete-time competing transitions, we derive exact transformations that map binary logistic models to valid competing transition probabilities, avoiding the need for continuous-time approximations. In simulations, our framework effectively recovers structured baseline and covariate effects while using the neural component to detect interaction patterns. We demonstrate the method using the Freddie Mac Single-Family Loan-Level Dataset, employing an out-of-time test design. Compared with a structured generalised additive benchmark, the semi-structured model provides modest but consistent gains in discrimination across the earliest prediction spans, while maintaining similar Brier scores. Adding macroeconomic indicators provides limited incremental benefit in this out-of-time evaluation and does not materially change the estimated borrower-, loan-, or duration-driven effects. Overall, semi-structured multi-state modelling offers a practical compromise between transparent effect estimates and flexible pattern learning, with potential applications beyond credit-transition forecasting.

stat.AP↗

arXiv2026

Scalable Learning of Multivariate Distributions via Coresets

Efficient and scalable non-parametric or semi-parametric regression analysis and density estimation are of crucial importance to the fields of statistics and machine learning. However, available methods are limited in their ability to handle large-scale data. We address this issue by developing a novel coreset construction for multivariate conditional transformation models (MCTMs) to enhance their scalability and training efficiency. To the best of our knowledge, these are the first coresets for semi-parametric distributional models. Our approach yields substantial data reduction via importance sampling. It ensures with high probability that the log-likelihood remains within multiplicative error bounds of $(1\pm\varepsilon)$ and thereby maintains statistical model accuracy. Compared to conventional full-parametric models, where coresets have been incorporated before, our semi-parametric approach exhibits enhanced adaptability, particularly in scenarios where complex distributions and non-linear relationships are present, but not fully understood. To address numerical problems associated with normalizing logarithmic terms, we follow a geometric approximation based on the convex hull of input data. This ensures feasible, stable, and accurate inference in scenarios involving large amounts of data. Numerical experiments demonstrate substantially improved computational efficiency when handling large and complex datasets, thus laying the foundation for a broad range of applications within the statistics and machine learning communities.

cs.LG↗

arXiv2026

Scalable and Robust Spatial Prediction via Multi-Resolution Ensembles of Predictive Processes

Gaussian processes provide a flexible framework for spatial prediction, but their computational cost limits applicability to large-scale data with large sample size $n$. Predictive processes (PPs), a popular low-rank approximation, mitigate this burden by projecting the original process onto a reduced set of $m\ll n$ inducing points. However, existing theory requires $m$ to grow with $n$, creating a trade-off between accuracy and computational efficiency. We address this challenge by introducing an ensemble of PPs based on spatial partitioning, and propose a novel partitioning and patching scheme with desirable properties. By generalizing the convergence results of PPs, it becomes possible to explicitly balance scalability and accuracy: increasing the number of ensemble components slows down the convergence but substantially improves computational efficiency. We further show theoretically that, despite the limited approximation accuracy of PPs with fixed $m$, they are asymptotically robust to data contamination. Motivated by this insight, we finally introduce a multi-resolution ensemble that combines PPs with fixed $m$ with multiple ensembles defined over possibly overlapping coarse to fine partitions. Simulations and large-scale geostatistical applications demonstrate that our approach delivers accurate, robust predictions with computational gains, providing a practical and broadly applicable solution for spatial prediction.

stat.ME↗

arXiv2026

Depth as Prior Knowledge for Object Detection

Detecting small and distant objects remains challenging for object detectors due to scale variation, low resolution, and background clutter. Safety-critical applications require reliable detection of these objects for safe planning. Depth information can improve detection, but existing approaches require complex, model-specific architectural modifications. We provide a theoretical analysis followed by an empirical investigation of the depth-detection relationship. Together, they explain how depth causes systematic performance degradation and why depth-informed supervision mitigates it. We introduce DepthPrior, a framework that uses depth as prior knowledge rather than as a fused feature, providing comparable benefits without modifying detector architectures. DepthPrior consists of Depth-Based Loss Weighting (DLW) and Depth-Based Loss Stratification (DLS) during training, and Depth-Aware Confidence Thresholding (DCT) during inference. The only overhead is the initial cost of depth estimation. Experiments across four benchmarks (KITTI, MS COCO, VisDrone, SUN RGB-D) and two detectors (YOLOv11, EfficientDet) demonstrate the effectiveness of DepthPrior, achieving up to +9% mAP$_S$ and +7% mAR$_S$ for small objects, with inference recovery rates as high as 95:1 (true vs. false detections). DepthPrior offers these benefits without additional sensors, architectural changes, or performance costs. Code is available at https://github.com/mos-ks/DepthPrior.

cs.CV↗

arXiv2026

Copula-based models for spatially dependent cylindrical data

Cylindrical data frequently arise across various scientific disciplines, including meteorology (e.g., wind direction and speed), oceanography (e.g., marine current direction and speed or wave heights), ecology (e.g., telemetry), and medicine (e.g., seasonality and intensity in disease onset). Such data often occur as spatially correlated series of intensities and angles, thereby representing dependent bivariate response vectors of linear and circular components. To accommodate both the circular-linear dependence and spatial autocorrelation, while remaining flexible in marginal specifications, copula-based models for cylindrical data have been developed in the literature. However, existing approaches typically treat the copula parameters as constants unrelated to covariates, and regression specifications for marginal distributions are frequently restricted to linear predictors, thereby ignoring spatial correlation. In this work, we propose a structured additive conditional copula regression model for cylindrical data. The circular component is modeled using a wrapped Gaussian process, and the linear component follows a distributional regression model. Both components allow for the inclusion of linear covariate effects. Furthermore, by leveraging the empirical equivalence between Gaussian random fields (GRFs) and Gaussian Markov random fields, our approach avoids the computational burden typically associated with GRFs, while simultaneously allowing for non-stationarity in the covariance structure. Posterior estimation is performed via Markov chain Monte Carlo simulation. We evaluate the proposed model in a simulation study and subsequently in an analysis of wind directions and speed in Germany.

stat.ME↗

arXiv2026

Bayesian Additive Regression Tree Copula Processes for Scalable Distributional Prediction

We show how to construct the implied copula process of response values from a Bayesian additive regression tree (BART) model with prior on the leaf node variances. This copula process, defined on the covariate space, can be paired with any marginal distribution for the dependent variable to construct a flexible distributional BART model. Bayesian inference is performed via Markov chain Monte Carlo on an augmented posterior, where we show that key sampling steps can be realized as those of Chipman et al. (2010), preserving scalability and computational efficiency even though the copula process is high dimensional. The posterior predictive distribution from the copula process model is derived in closed form as the push-forward of the posterior predictive distribution of the underlying BART model with an optimal transport map. Under suitable conditions, we establish posterior consistency for the regression function and posterior means and prove convergence in distribution of the predictive process and conditional expectation. Simulation studies demonstrate improved accuracy of distributional predictions compared to the original BART model and leading benchmarks. Applications to five real datasets with 506 to 515,345 observations and 8 to 90 covariates further highlight the efficacy and scalability of our proposed BART copula process model.

stat.ME↗

arXiv2025

BaGGLS: A Bayesian Shrinkage Framework for Interpretable Modeling of Interactions in High-Dimensional Biological Data

Biological data sets are often high-dimensional, noisy, and governed by complex interactions among sparse signals. This poses major challenges for interpretability and reliable feature selection. Tasks such as identifying motif interactions in genomics exemplify these difficulties, as only a small subset of biologically relevant features (e.g., motifs) are typically active, and their effects are often non-linear and context-dependent. While statistical approaches often result in more interpretable models, deep learning models have proven effective in modeling complex interactions and prediction accuracy, yet their black-box nature limits interpretability. We introduce BaGGLS, a flexible and interpretable probabilistic binary regression model designed for high-dimensional biological inference involving feature interactions. BaGGLS incorporates a Bayesian group global-local shrinkage prior, aligned with the group structure introduced by interaction terms. This prior encourages sparsity while retaining interpretability, helping to isolate meaningful signals and suppress noise. To enable scalable inference, we employ a partially factorized variational approximation that captures posterior skewness and supports efficient learning even in large feature spaces. In extensive simulations, we can show that BaGGLS outperforms the other methods with regard to interaction detection and is many times faster than MCMC sampling under the horseshoe prior. We also demonstrate the usefulness of BaGGLS in the context of interaction discovery from motif scanner outputs and noisy attribution scores from deep learning models. This shows that BaGGLS is a promising approach for uncovering biologically relevant interaction patterns, with potential applicability across a range of high-dimensional tasks in computational biology.

stat.ME↗