Ranked Activation Shift for Post-Hoc Out-of-Distribution Detection

2026-03-27

arXiv.org

AI Summary

Existing out-of-distribution (OOD) detection methods exhibit inconsistent performance across datasets and model architectures, particularly suffering significant degradation when activations in the penultimate layer are not rectified. To address this limitation, this work proposes RAS, a hyperparameter-free, plug-in OOD detection method that replaces sorted activation magnitudes with a fixed in-distribution reference activation spectrum. RAS makes no assumptions about the form of the activation function and enhances discriminability by leveraging both suppression and excitation of activation variations. Extensive experiments demonstrate that RAS consistently achieves strong OOD detection performance across diverse datasets and model architectures while preserving the original model’s classification accuracy.

Abstract

State-of-the-art post-hoc out-of-distribution detection methods rely on intermediate layer activation editing. However, they exhibit inconsistent performance across datasets and models. We show that this instability is driven by differences in the activation distributions, and identify a failure mode of scaling-based methods that arises when penultimate layer activations are not rectified. Motivated by this analysis, we propose \ours, a hyperparameter-free post-hoc method that replaces sorted activation magnitudes with a fixed in-distribution reference profile. Our simple plug-and-play method shows strong and consistent performance across datasets and architectures without assumptions on the penultimate layer activation function, and without requiring any hyperparameter tuning, while preserving in-distribution classification accuracy by construction. We further analyze what drives the improvement, showing that both inhibiting and exciting activation shifts independently contribute to better out-of-distribution discrimination.

Problem

• out-of-distribution detection
• activation distribution
• post-hoc method
• model instability
• penultimate layer

Innovation

• Ranked Activation Shift
• out-of-distribution detection
• activation editing
• hyperparameter-free
• post-hoc method