Bayesian workflow diagnostics outperform unit tests for detecting and repairing statistically misspecified LLM-generated probabilistic programs across benchmarks and real generation tasks.
Journal of statistical software , volume=
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
years
2026 5representative citing papers
Adaptively learning the auxiliary ellipse distribution in elliptical slice sampling gives a gradient-free sampler that is ergodic under stated assumptions and empirically competitive with HMC and adaptive random walks on challenging posteriors.
A Bayesian hyperbolic latent space model with an inferred temperature parameter outperforms fixed-temperature and Euclidean alternatives in network reconstruction on simulated and real data.
Embedding selection mechanisms into generative simulators enables amortized Bayesian inference to produce debiased, well-calibrated posteriors without tractable likelihoods.
Measurement error in latent confounders produces biased ATE estimates and miscalibrated intervals under conventional adjustment; a Bayesian joint model of measurement, treatment, and outcome is proposed to correct it.
citing papers explorer
-
Calibration, Not Compilation: Detecting and Repairing Misspecified Probabilistic Programs Written by Language Models
Bayesian workflow diagnostics outperform unit tests for detecting and repairing statistically misspecified LLM-generated probabilistic programs across benchmarks and real generation tasks.
-
Adaptive Generalized Elliptical Slice Sampling
Adaptively learning the auxiliary ellipse distribution in elliptical slice sampling gives a gradient-free sampler that is ergodic under stated assumptions and empirically competitive with HMC and adaptive random walks on challenging posteriors.
-
Hyperbolic Latent Space Models for Network Embedding: Model Specification and Bayesian Inference
A Bayesian hyperbolic latent space model with an inferred temperature parameter outperforms fixed-temperature and Euclidean alternatives in network reconstruction on simulated and real data.
-
Overcoming Selection Bias in Statistical Studies With Amortized Bayesian Inference
Embedding selection mechanisms into generative simulators enables amortized Bayesian inference to produce debiased, well-calibrated posteriors without tractable likelihoods.
-
Measurement Induced Confounding
Measurement error in latent confounders produces biased ATE estimates and miscalibrated intervals under conventional adjustment; a Bayesian joint model of measurement, treatment, and outcome is proposed to correct it.