DAVIS : OOD Detection via Dominant Activations and Variance for Increased Separation

Abid Hassan, Tuan Ngo, Saad Shafiq, Nenad Medvidovic · 2026

Read DAVIS : OOD Detection via Dominant Activations and Variance for Increased Separation by Abid Hassan, Tuan Ngo, Saad Shafiq, Nenad Medvidovic (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-22703.txt.

Description

Abid Hassan, Tuan Ngo, Saad Shafiq, and Nenad Medvidovic built DAVIS, a detector for inputs unlike anything a model was trained on. Out-of-distribution detection is a key guard for real-world deployment. Their 2026 preprint notes that most post-hoc tools read a pooled feature summary, which throws away rich detail from the activation maps.

DAVIS uses dominant activations and their variance to widen the gap between known and strange inputs. A wider gap means fewer strangers waved through. The work is a guard-rail upgrade. It costs no retraining of the base model. That ease is part of its appeal.

Catalog record

CREATORAbid Hassan, Tuan Ngo, Saad Shafiq, Nenad Medvidovic
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE159 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-22703

Browse the full library · About the Tower of Babel