Shaping Parameter Contribution Patterns for Out-of-Distribution Detection

TL;DR

This paper introduces SPCP, a training method that encourages classifiers to use more parameters for decisions, improving out-of-distribution detection by reducing overconfidence caused by sparse parameter reliance.

Contribution

We propose SPCP, a novel training technique that promotes dense parameter contribution patterns, enhancing OOD detection robustness without sacrificing in-distribution accuracy.

Findings

SPCP improves OOD detection performance across various benchmarks.

It reduces overconfidence by encouraging broader parameter utilization.

Experimental results show enhanced robustness without harming ID accuracy.

Abstract

Out-of-distribution (OOD) detection is a well-known challenge due to deep models often producing overconfident. In this paper, we reveal a key insight that trained classifiers tend to rely on sparse parameter contribution patterns, meaning that only a few dominant parameters drive predictions. This brittleness can be exploited by OOD inputs that anomalously trigger these parameters, resulting in overconfident predictions. To address this issue, we propose a simple yet effective method called Shaping Parameter Contribution Patterns (SPCP), which enhances OOD detection robustness by encouraging the classifier to learn boundary-oriented dense contribution patterns. Specifically, SPCP operates during training by rectifying excessively high parameter contributions based on a dynamically estimated threshold. This mechanism promotes the classifier to rely on a broader set of parameters for decision-making, thereby reducing the risk of overconfident predictions caused by anomalously triggered parameters, while preserving in-distribution (ID) performance. Extensive experiments under various OOD detection setups verify the effectiveness of SPCP.