ConformaDecompose: Explaining Uncertainty via Calibration Localization

TL;DR

ConformaDecompose introduces a framework for analyzing and explaining the sources of uncertainty in conformal prediction intervals at the instance level, enhancing interpretability.

Contribution

It provides a diagnostic method to localize calibration support and explain the reducibility of epistemic uncertainty in regression tasks.

Findings

Absolute reducible uncertainty aligns with epistemic proxies.

Relative contribution of uncertainty varies by task.

The approach enhances interpretability without changing the predictor or coverage.

Abstract

Conformal Prediction provides distribution-free prediction intervals with guaranteed coverage, but its reliance on a single global calibration threshold obscures the sources of uncertainty at the instance level. In particular, it conflates irreducible noise with uncertainty induced by heterogeneous training data (aleatoric), model limitations, or calibration mismatch (epistemic), offering little insight into why an interval is wide or whether it could be reduced. We introduce an uncertainty-aware explainability framework that analyses the reducibility of calibration-induced epistemic conformal uncertainty via progressive calibration localisation for regression tasks. The approach is diagnostic rather than causal: it does not estimate true aleatoric or epistemic uncertainty, but explains how conformal intervals contract and stabilise as calibration support is localised around a test instance. Across benchmarks and real-world data, absolute reducible uncertainty aligns with epistemic proxies, while its relative contribution varies by task, revealing regimes hidden by interval width. This instance-level view complements conformal uncertainty, enhancing interpretability without altering the predictor or coverage.