Saved in:
Bibliographic Details
Main Authors: Zhu, Fei, Zhang, Zhaoxiang
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2504.14545
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910915236986880
author Zhu, Fei
Zhang, Zhaoxiang
author_facet Zhu, Fei
Zhang, Zhaoxiang
contents Reliable prediction is an essential requirement for deep neural models that are deployed in open environments, where both covariate and semantic out-of-distribution (OOD) data arise naturally. In practice, to make safe decisions, a reliable model should accept correctly recognized inputs while rejecting both those misclassified covariate-shifted and semantic-shifted examples. Besides, considering the potential existing trade-off between rejecting different failure cases, more convenient, controllable, and flexible failure detection approaches are needed. To meet the above requirements, we propose a simple failure detection framework to unify and facilitate classification with rejection under both covariate and semantic shifts. Our key insight is that by separating and consolidating failure-specific reliability knowledge with low-rank adapters and then integrating them, we can enhance the failure detection ability effectively and flexibly. Extensive experiments demonstrate the superiority of our framework.
format Preprint
id arxiv_https___arxiv_org_abs_2504_14545
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle TrustLoRA: Low-Rank Adaptation for Failure Detection under Out-of-distribution Data
Zhu, Fei
Zhang, Zhaoxiang
Machine Learning
Reliable prediction is an essential requirement for deep neural models that are deployed in open environments, where both covariate and semantic out-of-distribution (OOD) data arise naturally. In practice, to make safe decisions, a reliable model should accept correctly recognized inputs while rejecting both those misclassified covariate-shifted and semantic-shifted examples. Besides, considering the potential existing trade-off between rejecting different failure cases, more convenient, controllable, and flexible failure detection approaches are needed. To meet the above requirements, we propose a simple failure detection framework to unify and facilitate classification with rejection under both covariate and semantic shifts. Our key insight is that by separating and consolidating failure-specific reliability knowledge with low-rank adapters and then integrating them, we can enhance the failure detection ability effectively and flexibly. Extensive experiments demonstrate the superiority of our framework.
title TrustLoRA: Low-Rank Adaptation for Failure Detection under Out-of-distribution Data
topic Machine Learning
url https://arxiv.org/abs/2504.14545