Towards a Certificate of Trust: Task-Aware OOD Detection for Scientific AI

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Raonić, Bogdan, Mishra, Siddhartha, Lanthaler, Samuel
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909032207351808
author Raonić, Bogdan
Mishra, Siddhartha
Lanthaler, Samuel
author_facet Raonić, Bogdan
Mishra, Siddhartha
Lanthaler, Samuel
contents Data-driven models are increasingly adopted in critical scientific fields like weather forecasting and fluid dynamics. These methods can fail on out-of-distribution (OOD) data, but detecting such failures in regression tasks is an open challenge. We propose a new OOD detection method based on estimating joint likelihoods using a score-based diffusion model. This approach considers not just the input but also the regression model's prediction, providing a task-aware reliability score. Across numerous scientific datasets, including PDE datasets, satellite imagery and brain tumor segmentation, we show that this likelihood strongly correlates with prediction error. Our work provides a foundational step towards building a verifiable 'certificate of trust', thereby offering a practical tool for assessing the trustworthiness of AI-based scientific predictions. Our code is publicly available at https://github.com/bogdanraonic3/OOD_Detection_ScientificML
format Preprint
id arxiv_https___arxiv_org_abs_2509_25080
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Towards a Certificate of Trust: Task-Aware OOD Detection for Scientific AI
Raonić, Bogdan
Mishra, Siddhartha
Lanthaler, Samuel
Machine Learning
Data-driven models are increasingly adopted in critical scientific fields like weather forecasting and fluid dynamics. These methods can fail on out-of-distribution (OOD) data, but detecting such failures in regression tasks is an open challenge. We propose a new OOD detection method based on estimating joint likelihoods using a score-based diffusion model. This approach considers not just the input but also the regression model's prediction, providing a task-aware reliability score. Across numerous scientific datasets, including PDE datasets, satellite imagery and brain tumor segmentation, we show that this likelihood strongly correlates with prediction error. Our work provides a foundational step towards building a verifiable 'certificate of trust', thereby offering a practical tool for assessing the trustworthiness of AI-based scientific predictions. Our code is publicly available at https://github.com/bogdanraonic3/OOD_Detection_ScientificML
title Towards a Certificate of Trust: Task-Aware OOD Detection for Scientific AI
topic Machine Learning
url https://arxiv.org/abs/2509.25080