archive
Every paper Pith has read. Search by title, abstract, or pith.
6894 papers in stat · page 1
-
Simple threshold monitor matches advanced LLM safety checks
Online Safety Monitoring for LLMs
-
CESM low-energy matches contrast at 0.874 patient AUC
Masked complex non-decimated wavelet features for patient-level classification of contrast-enhanced mammography
-
The paper develops design-based inference for experiments that randomly assign…
Inference for Group Interaction Experiments
-
LLM personas split into frame-stable aggregates and frame-sensitive geometry
The Dual Nature of LLM Persona: Aggregated Tendencies and Frame-Dependent Geometry
-
Bayesian and quasi-Bayesian estimates merge for Poisson decisions
Merging of Bayes and quasi-Bayes empirical Bayes procedures for Poisson compound decisions
-
CAP matches empirical risk at first order and removes second-order bias
Cross-Audit Projection for Model Risk Prediction
-
iDiD estimator lets time-constant direct effects count as valid instruments
Instrumented difference-in-differences under case-control sampling
-
EVPI extensions show global validation of ADNEX model is complete
Value-of-Information Analysis for External Validation of Risk Prediction Models in Multicenter Studies and Systematic Reviews
-
Multipliers create nondegenerate tests for Fréchet regression
MATCH: Multiplier-Assisted Tests for Conditional Hypotheses in Non-Euclidean Data
-
Joint distributions of sample statistics power new GOF tests
Goodness of Fit Tests Based on Joint Densities of Multiple Sample Statistics
-
AEW achieves T log(M)/(n+1) excess risk in expectation
Aggregation with Exponential Weights is Optimal in Expectation
-
Additive MLP-GNN separates chemical and structural solubility drivers
An Additive MLP-GNN Framework for Characterizing Chemical and Structural Contributions to Aqueous Solubility
-
Policy-coupled coverage optimizes counterfactual prediction sets
Prediction Sets for Counterfactual Decisions: Coverage, Optimality, and Conformal Prediction
-
Weighted tilt restores coverage for censored label shift
Conformal Bayes for Two-Sided Censored Gaussian Regression under Label Shift
-
Weaker matrix condition extends simplex volume theorems to AR(1) models
A note on "The volume of random simplices from elliptical distributions in high dimension"
-
Quaternion wavelets classify breast histology into four classes
Quaternion Nondecimated Wavelet Descriptors for Multiclass Breast Histology Classification
-
Conformal prediction flags 15 percent of samples but detects zero traction incidents
Predictive Conformal Slip Monitoring: An Empirical Evaluation of Rolling Split Conformal Prediction for Pre-Incident Traction Loss Detection
-
Expert portfolio detects PDE model errors missed by residuals
Sequential Structure-Sensitive Residual Diagnostics for PDE Inverse Problems
-
GIV consistent at √T only when few units dominate aggregate
Granular Instrumental Variables in Large Panels: Identification and Inference Across Strong, Nearly Weak, and Weak GIV
-
Orthogonal arrays and difference schemes create larger grouped arrays
Grouped Orthogonal Arrays from Orthogonal Arrays and Difference Schemes
-
Geometric graphs indistinguishable from random above (nh(p))^3 dims
Resolution of the Detection Threshold Conjecture for Random Geometric Graphs in the $d>n$ Regime
-
Shallow network optimum recovered by one linear solve
Born Discrete, Made Smooth: Variational Formulation of Shallow Neural Networks
-
Probit BKMR fits in bkmr rarely converge
Convergence fragility in probit Bayesian kernel machine regression implemented in the bkmr R package for binary-outcome environmental mixture analyses: a simulation study
-
Plausibility enables exact inference in general parametric families
Plausibility: Exact inference in R
-
Moment method selects random effects consistently in mixed models
Moment-Based Selection of Multiresponse Linear Mixed-Effects Models
-
IRT model extracts rider skill and condition difficulty from binary outcomes
Inverse Suitability: Identifying Condition Difficulty and Rider Skill from Behavioural Outcomes via Continuous-Item Response Theory
-
Autorelevance function recovers lag structure in time series forecasts
Autorelevance function and other feature relevance measures for univariate time series
-
K-means centers from MCAR data converge at √n rate
Statistical Properties of $k$-means Clustering for Data Missing Completely at Random
-
Perturbation theory transfers sup-norm rates to functional principal components
Transferring supremum-norm rates and weak convergence of covariance kernel estimators to functional principal components
-
Closed-form wrapped Gaussians replace Laplace for posterior approximation
Beyond Laplace: Closed-form wrapped Gaussian posterior approximations on statistical manifolds
-
Variational estimator beats spectral for density ratios with abundant data
Regularized Variational and Spectral Log-Density-Ratio Estimation in the Gaussian Location Model
-
Calibration lets multimodal predictors borrow across missingness patterns
Pattern-Calibrated Multimodal Prediction under Blockwise Missingness
-
Role projections lift directional accuracy on semantic benchmarks
Role-Aware Neural Convex Divergence Heads for Asymmetric Representation Learning
-
Point source restores identifiability of spatial dynamics from snapshots
Identifiability Limits of Physics-Informed Inference for Spatial Stochastic Dynamics from Static Snapshots
-
Method brings full Bayesian inference to RL without likelihoods
Full Bayesian Reinforcement Learning via LF-IBIS
-
Glicko-2 extended with margin and draw models for football
An Adaptive Glicko-2 Rating Framework for Probabilistic Football Forecasting and Season Simulation
-
Transportability estimates effect changes from modifier prevalence shifts
From Subgroups to Population Composition: A Transportability Approach to Effect Heterogeneity
-
-
Spike-and-slab prior recovers multimorbidity clusters in EHR data
Continuous-Time Bayesian Networks with Structured Shrinkage Priors for Modelling Multimorbidity Trajectories in Large-Scale Electronic Health Records
-
Soliton dynamics recovered from scattering data without equations
Learning Effective Soliton Dynamics from Scattering Data
-
Kernel norm spots PV faults at 99% accuracy without labels
An unsupervised kernel norm monitoring for fault detection in a time series photovoltaic system
-
AI agents reproduce 72% of human ideological gaps in data analysis
The Agentic Garden of Forking Paths
-
-
Three-term law recovers optimal batch size from suboptimal runs
How to Allocate Your Tokens? Scaling Laws with Training Steps and Batch Size
-
Parameter uncertainty reduces epidemic sensitivity and yields conservative policies
Sensitivity Analysis and Optimization of Stochastic Epidemic Models under Parameter Uncertainty
-
20-50 sampled points per region lift spatial scan power
Sampling for Region-Aggregated Spatial Scan Statistics
-
Conditional forests rank 3rd-4th as feature selectors
Conditional Inference Trees and Forests for Feature Selection
-
Score test distinguishes non-nested cure-fraction survival models
J- and MJ-Type Tests for Non-Nested Parametric Survival Models with a Cure Fraction: A Score Test Approach
-
Motif signatures distinguish latent positions where degrees match
Beyond Degree: Rooted Motif Signatures for Latent Position Identifiability in Graphon Models
-
One narrative links approximation to emergence in deep learning
From Approximation to Emergence: A Theory of Deep Learning