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

TL;DR

This paper introduces , a hyperparameter-free post-hoc method for out-of-distribution detection that uses a fixed reference profile to improve consistency across datasets and models.

Contribution

It proposes a novel activation shift method that is simple, hyperparameter-free, and improves out-of-distribution detection performance without affecting in-distribution accuracy.

Findings

 outperforms existing methods across various datasets and architectures.

The method is hyperparameter-free and does not require tuning.

Activation shifts independently contribute to improved detection.

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.