Low-dimensional embeddings are widely used as visual summaries of high-dimensional data and to enable downstream scientific discoveries. Yet, popular nonlinear dimension reduction methods, such as t-SNE and UMAP, are often selected based on visual appeal alone and without rigorous quantitative validation. A major reason is that manifold embeddings typically do not provide an out-of-sample map nor an inverse back to the original feature space; this makes held-out validation, the gold standard in supervised learning, all but impossible. To address these challenges, we develop a novel framework, MEDAL (Manifold Embedding Distillation via Autoencoder Learning), which distills a fitted manifold embedding into a reusable encoder--decoder model. MEDAL trains a constrained autoencoder whose bottleneck exactly matches any teacher embedding while the decoder reconstructs the original input; this yields an explicit map for new samples, an approximate inverse, and a pointwise reconstruction-based measure of distortion in the manifold space. This converts static manifold embeddings into models that can be evaluated on held-out data, enabling quantitative validation including comparing different dimension reduction methods as well as hyperparameter tuning. Across multiple benchmark and scientific case studies, we show that MEDAL enables held-out validation to determine optimal manifold embeddings and hyperparameters, reveals biologically coherent regions that are difficult to preserve in two dimensional embeddings, and detects distribution shift when new samples are mapped into a fixed reference manifold. MEDAL provides a general validation wrapper to any existing dimension reduction technique that will improve the rigor and

We prove that no reinforcement learning policy with confidence-gated autonomy can simultaneously achieve maximum helpfulness, optimal calibration, and full autonomy under rational oversight, whenever some tasks exceed the agent's reliable competence: the Behavioral Credibility Trilemma. The impossibility is geometric -- adding any non-affine autonomy incentive to a strictly proper scoring rule destroys strict properness, so an agent rewarded for both calibrated confidence and autonomous action systematically inflates its reported confidence on tasks below the principal's approval threshold. The Behavioral Perturbation Lemma quantifies the inflation (scaling as $w_A/(2 w_C)$ for the Brier score) and shows detection requires $Ω(1/Δ^2)$ observations. We prove the principal's optimal oversight rule is necessarily non-affine, making the impossibility unconditional and optimizer-independent across log-concave-density policy families. We formalize the Confidence-Gated Decision Problem, map existing methods onto the trilemma, and identify two constructive resolution pathways (commitment, domain separation). A 540-configuration Best-of-N experiment tests five pre-registered hypotheses, all strongly confirmed (effect sizes $d = 1.10$ to $5.32$), and adds a descriptive analysis of the achievable-$(H, C, A)$ surface geometry showing a plateau-truncated frontier consistent with the predicted inflation saturation.

In the era of precision medicine, genome-wide epigenetic modifications offer rich data that could inform risk prediction. However, these data are high-dimensional and exhibit complex dependence structures, which makes it difficult to jointly model them with low-dimensional covariates when the goal is to obtain interpretable effect estimates for covariate adjustment. Standard Bayesian additive regression trees (BART) provide strong predictive performance but treat all predictors uniformly within the tree ensemble, obscuring the contributions of significant covariates and complicating variable selection in high-dimensional settings. We propose a semi-parametric BART model (spBART) that addresses this limitation by modeling low-dimensional covariates through a parametric component with interpretable coefficients, while capturing complex nonlinear associations among high-dimensional predictors through the tree ensemble. To perform stable variable selection, we develop a cross-validation-based procedure that aggregates posterior inclusion probabilities across folds and applies Bayesian false discovery rate control. We apply the proposed method to a pooled case--control analysis of high-dimensional genome-wide 5-hydroxymethylcytosine profiles derived from circulating cell-free DNA in two multiple myeloma studies ($N = 869$). The approach identifies a parsimonious set of candidate loci and achieves strong out-of-sample discrimination (AUC $= 0.96$) in a held-out validation set. Overall, spBART provides a unified framework for combining interpretable covariate inference with flexible modeling and variable selection in high-dimensional biomedical studies.

Large language models (LLMs) offer a scalable mechanism to elicit domain-informed prior information for high-dimensional variable selection. However, existing methods such as LLM-Lasso are sensitive to weight quality, with performance degrading substantially when LLM-generated weights are inaccurate. To address this challenge, we first introduce a framework for quantifying the quality of LLM-generated weights, enabling rigorous evaluation of LLM-informed methods across varying weight regimes. We then propose the LLM Sparsity Prior (LSP), which integrates LLM-generated weights into the prior inclusion probabilities of Spike-and-Slab and Spike-and-Slab Lasso models via two interpretable hyperparameters governing global sparsity and weight concentration. Hierarchical hyperpriors on these parameters allow the model to dynamically discount uninformative or misleading weights, improving robustness without sacrificing gains when weights are accurate. Finally, we develop principled prompt engineering strategies and validate the method on a private medical dataset studying Acute Kidney Injury. LSP improves prediction accuracy and identifies clinically relevant features missed by the baselines, with robustness to prompt variation and particular effectiveness in low-data regimes.

Logged advertising auctions make offline reserve-price evaluation attractive but risky. Replay tables can identify policies with large apparent yield gains, yet they can also hide weak threshold support, multiple-comparison effects, subgroup harm, and bidder-response uncertainty. Existing replay and off-policy evaluation methods estimate or rank policy values, but they do not directly answer the operational question of whether the available evidence is strong enough to justify validation. This paper develops a support-aware offline decision framework for reserve-policy selection. Rather than outputting a single point-estimate winner, the framework converts logged evidence into a conservative decision object consisting of certified policies, statistically dominated alternatives, and unresolved candidates requiring further validation. The main theoretical result gives a unified finite-catalog guarantee showing that, under simultaneous uncertainty control and conservative support gates, the framework preserves the best gate-passing policy while eliminating only policies with certified regret. Supporting results characterize support-localized replay generalization, establish information-theoretic threshold-resolution limits, and quantify when heterogeneous bidder response can overturn localized replay rankings. Experiments on iPinYou real-time-bidding logs show that the leading reserve rule achieves a 47.66% replay lift in season two, a 40.71% simultaneous lower-bound lift, and a 43.87% frozen out-of-time replay lift in season three. The framework reduces a 19-policy catalog to a two-policy validation shortlist while certifying non-harm across 44 advertiser, exchange, and region segments. The results support the central claim that offline reserve-policy evaluation should produce certified validation decisions rather than point-estimate rankings alone.

Interpretable text representations should expose coordinates that are not only predictive, but also meaningful enough for independent auditors to apply. Existing discriminative representations often use anonymous embedding directions, while concept-bottleneck and LLM-assisted methods attach natural-language names to features without ensuring that those definitions are reproducible or distinct from the target label. We propose an operational criterion for interpretable discriminative text representations: each coordinate should satisfy conceptual clarity, measured by chance-adjusted agreement between independent annotators applying the feature definition, and label disentanglement, meaning the feature should not merely paraphrase the prediction target. We instantiate this criterion in LLM-assisted Feature Discovery (LFD), an iterative method that proposes lexical and semantic features from contrastive outcome-opposed text pairs, screens candidates using cross-LLM Cohen's $κ$, and selects features by residual held-out predictive gain. A stylized analysis connects the $κ$ screen to a per-feature annotation-noise bound, formalizing agreement as a reliability check. Across ten text-classification tasks spanning seven corpora, LFD matches the predictive performance of a strong text bottleneck baseline while producing substantially clearer and less label-entangled features. Human audits with 232 raters show that LFD features achieve higher human--human and human--LLM agreement than baseline concepts, and raters consistently judge them as less label-leaking. These results suggest that agreement-tested, label-disentangled coordinates provide a practical auditability standard for interpretable text classification.

Sequential prediction is challenging in regimes of delayed disambiguation, where early observations are ambiguous and multiple latent explanations remain plausible until sufficient evidence accumulates. Standard approaches based on marginal inference struggle in this setting, either collapsing uncertainty prematurely or failing to recover once informative evidence arrives. We introduce EviTrack, a test-time inference framework that operates over latent trajectories rather than marginal states. EviTrack maintains a set of competing trajectory hypotheses and applies evidence- and likelihood-ratio-based selection to delay commitment until supported by data, drawing inspiration from hypothesis management in multiple hypothesis tracking and track-before-detect. To evaluate this setting, we construct a controlled synthetic benchmark with known latent ground truth that explicitly exhibits delayed disambiguation. At matched inference budget, EviTrack substantially outperforms sampling-based baselines, achieving faster post-disambiguation recovery. These results show that, in delayed disambiguation regimes, moderate trajectory-level selection is more effective than increasing sampling coverage, highlighting selection over sampling as a key principle for reliable sequential inference.

We propose PathBoost, a gradient tree boosting method for graph-level classification and regression that learns discriminative path-based features directly from the input graph structure. Building on a previous work, which was tailored to a specific chemistry application, PathBoost introduces three key extensions: (i) adaptation to binary classification through gradient boosting with a logistic loss, (ii) incorporation of multiple node and edge attributes into the path feature space via a prefix-based decomposition, and (iii) automatic anchor node selection based on categorical attribute diversity, eliminating the need for the user to specify the starting point of the considered path features. We compared PathBoost to graph neural networks and graph kernel approaches on several benchmark datasets, obtaining better results in half of them, and comparable results in the rest. PathBoost shows better performances on graphs with larger average node counts. Overall, the results demonstrate that path-based boosting methods can be competitive with more complex black-box approaches.

Adaptive prompt and program search makes LLM evaluation selection-sensitive. Once benchmark items are reused inside tuning, the observed winner's score need not estimate the fresh-data performance of the full tune-then-deploy procedure. We study inference for this procedure-level target under explicit tuning budgets. We propose SIREN, a selection-aware repeated-split reporting protocol that freezes the post-search shortlist, separates splitwise selection from held-out evaluation, and uses an item-level Gaussian multiplier bootstrap for uncertainty quantification. In a fixed-shortlist regime with smooth stabilized selection, the estimator admits a first-order item-level representation, and the bootstrap yields valid simultaneous inference on a finite budget grid. This supports confidence intervals for procedureperformance curves and pre-specified equal-budget and cross-budget comparisons. Controlled simulations and MMLU-Pro tuning experiments show that winnerbased reporting can be optimistic and can change deployment conclusions, while SIREN remains close to the finite-sample reporting target. Codes are available at https://github.com/jznmsl/siren.

Randomized controlled trials typically assume that prognostic covariates are known and available at no cost. In practice, obtaining high-dimensional pretreatment data is costly, forcing a trade-off between covariate-adaptive precision and a measurement budget. We introduce Dynamic Adaptive Rerandomization via Thompson Sampling (DARTS), which treats covariate acquisition as a sequential optimization problem embedded within a design-based causal inference task. A budgeted combinatorial Thompson sampler learns which covariates are most prognostic across successive batches; selected covariates then drive rerandomization and regression adjustment to reduce batch-level average treatment effect variance. Our primary theoretical contribution is a decoupling result: adaptive covariate selection based on past batches preserves batch-level randomization validity, and the cumulative inverse-variance weighted estimator achieves at least nominal asymptotic coverage. We further derive a Bayes risk bound for the acquisition layer that matches the minimax lower bound up to logarithmic factors. Empirically, DARTS systematically concentrates the budget on informative features, significantly closing the efficiency gap to oracle designs while maintaining strict inferential validity.