Shaping Parameter Contribution Patterns for Out-of-Distribution Detection
This work aims to improve the robustness of OOD detection for deep learning models, which is a critical problem for deploying AI systems in real-world scenarios where unexpected inputs are common.
This paper addresses the problem of overconfident predictions in Out-of-Distribution (OOD) detection by revealing that deep models rely on sparse parameter contribution patterns. They propose Shaping Parameter Contribution Patterns (SPCP), a method that encourages dense, boundary-oriented contribution patterns during training by rectifying high parameter contributions, thereby reducing overconfidence in OOD inputs.
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.