Adaptive Conformal Prediction for Image Regression Models with Application to an Inertial Confinement Fusion Emulator

Uncertainty quantification is critical in scientific machine learning, where black-box, image-based models are increasingly deployed in high-stakes settings.