English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Out-of-distribution (OOD) detection is an indispensable aspect of secure AIwhen deploying machine learning models in real-world applications. Previousparadigms either explore better scoring functions or utilize the knowledge ofoutliers to equip the models with the ability of OOD detection. However, few ofthem pay attention to the intrinsic OOD detection capability of the givenmodel. In this work, we generally discover the existence of an intermediatestage of a model trained on in-distribution (ID) data having higher OODdetection performance than that of its final stage across different settings,and further identify one critical data-level attribution to be learning withthe atypical samples. Based on such insights, we propose a novel method,Unleashing Mask, which aims to restore the OOD discriminative capabilities ofthe well-trained model with ID data. Our method utilizes a mask to figure outthe memorized atypical samples, and then finetune the model or prune it withthe introduced mask to forget them. Extensive experiments and analysisdemonstrate the effectiveness of our method. The code is available at:https://github.com/tmlr-group/Unleashing-Mask.

Unleashing Mask: Explore the Intrinsic Out-of-Distribution Detection Capability