CP4SBI: Local Conformal Calibration of Credible Sets in Simulation-Based Inference
This addresses the issue of unreliable uncertainty quantification for experimental scientists using SBI, though it appears incremental as it builds on existing conformal calibration methods.
The paper tackled the problem of miscalibrated posterior approximations in simulation-based inference (SBI), which cause credible regions to undercover true parameters, and developed CP4SBI, a model-agnostic conformal calibration framework that constructs credible sets with local Bayesian coverage, improving uncertainty quantification in experiments on SBI benchmarks.
Current experimental scientists have been increasingly relying on simulation-based inference (SBI) to invert complex non-linear models with intractable likelihoods. However, posterior approximations obtained with SBI are often miscalibrated, causing credible regions to undercover true parameters. We develop $\texttt{CP4SBI}$, a model-agnostic conformal calibration framework that constructs credible sets with local Bayesian coverage. Our two proposed variants, namely local calibration via regression trees and CDF-based calibration, enable finite-sample local coverage guarantees for any scoring function, including HPD, symmetric, and quantile-based regions. Experiments on widely used SBI benchmarks demonstrate that our approach improves the quality of uncertainty quantification for neural posterior estimators using both normalizing flows and score-diffusion modeling.