Beyond Prediction: Conformal Inference for Latent Distributional Parameters
Abstract
Many prediction problems seek to infer an unobserved, instance-specific parameter that governs the distribution of an observable response, even though the latent parameter is unavailable for both historical and future instances. We develop LatentCP, a prior-free conformal framework that constructs uncertainty sets for latent distributional parameters using only observed context--response pairs and a specified forward model. The method first constructs a conformal prediction set in the observable response space and then retains candidate latent parameters according to the probability their induced response distributions assign to that set. This inversion provides finite-sample marginal coverage without requiring latent calibration labels, a unique inverse mapping, or knowledge of the latent mixing distribution. Because latent-set efficiency depends nonmonotonically on the response-space miscoverage rate, we further introduce a multilevel procedure that aggregates normalized incompatibility scores across several response sets and selects the aggregation distribution using an independent tuning sample. Across synthetic experiments, LatentCP maintains nominal latent coverage under weak forward identification, observational nonidentifiability, and latent heterogeneity and multimodality, where empirical-Bayes, likelihood-based, and proxy-label conformal methods can substantially under-cover. On a California wildfire real dataset, it produces spatially adaptive uncertainty sets for latent fire intensity. Independent tuning improves efficiency, while multilevel aggregation provides additional gains when different response levels contain complementary information, without sacrificing validity.
Related
- Auditing Conformal Prediction for Language Models Under Inference-Time Model Shift
- Beyond Imputation: Mask-Adaptive Conformal Prediction via Tree Embeddings on General Missing Data Mechanisms
- Beyond Point Predictions: Distribution-Informed Prediction-Powered Inference
- Online Inference in Distributional Temporal-Difference Learning
- Orthogonal Inference for Conditional Z-Estimation without Cross-Fitting: A Distributional Nearest-Neighbor Approach
BibTeX
@misc{zheng2026beyond,
title = {{Beyond Prediction: Conformal Inference for Latent Distributional Parameters}},
author = {Minxing Zheng and Wenbin Zhou and Shixiang Zhu},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2608.03607},
eprint = {2608.03607},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2608.03607},
}
RIS
TY - UNPB
TI - Beyond Prediction: Conformal Inference for Latent Distributional Parameters
AU - Minxing Zheng
AU - Wenbin Zhou
AU - Shixiang Zhu
PY - 2026
DA - 2026/09/02
PB - arXiv
DO - 10.48550/arxiv.2608.03607
UR - https://arxiv.org/abs/2608.03607
AB - Many prediction problems seek to infer an unobserved, instance-specific parameter that governs the distribution of an observable response, even though the latent parameter is unavailable for both historical and future instances. We develop LatentCP, a prior-free conformal framework that constructs uncertainty sets for latent distributional parameters using only observed context--response pairs and a specified forward model. The method first constructs a conformal prediction set in the observable response space and then retains candidate latent parameters according to the probability their induced response distributions assign to that set. This inversion provides finite-sample marginal coverage without requiring latent calibration labels, a unique inverse mapping, or knowledge of the latent mixing distribution. Because latent-set efficiency depends nonmonotonically on the response-space miscoverage rate, we further introduce a multilevel procedure that aggregates normalized incompatibility scores across several response sets and selects the aggregation distribution using an independent tuning sample. Across synthetic experiments, LatentCP maintains nominal latent coverage under weak forward identification, observational nonidentifiability, and latent heterogeneity and multimodality, where empirical-Bayes, likelihood-based, and proxy-label conformal methods can substantially under-cover. On a California wildfire real dataset, it produces spatially adaptive uncertainty sets for latent fire intensity. Independent tuning improves efficiency, while multilevel aggregation provides additional gains when different response levels contain complementary information, without sacrificing validity.
ER -