Advancing Out-of-Distribution Detection via Local Neuroplasticity

2025-02-20

AI Summary

In real-world scenarios, distribution shifts (out-of-distribution, OOD) between training and test data undermine model reliability. To address this, we propose the first unsupervised, task-agnostic OOD detection paradigm grounded in the local neural plasticity of Kolmogorov–Arnold Networks (KANs). Leveraging KANs’ inherently interpretable activation structure, our method quantifies per-sample changes in neural activation patterns before and after training—using this discrepancy as the OOD score—without fine-tuning or auxiliary supervision. Evaluated across diverse image and medical benchmarks, our approach consistently outperforms existing state-of-the-art methods, demonstrating superior generalization, robustness to perturbations, and cross-task transferability. By eliminating reliance on distribution-specific assumptions or heavy architectural modifications, it establishes a novel, interpretable, lightweight, and distribution-agnostic pathway toward trustworthy machine learning.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

In the domain of machine learning, the assumption that training and test data share the same distribution is often violated in real-world scenarios, requiring effective out-of-distribution (OOD) detection. This paper presents a novel OOD detection method that leverages the unique local neuroplasticity property of Kolmogorov-Arnold Networks (KANs). Unlike traditional multilayer perceptrons, KANs exhibit local plasticity, allowing them to preserve learned information while adapting to new tasks. Our method compares the activation patterns of a trained KAN against its untrained counterpart to detect OOD samples. We validate our approach on benchmarks from image and medical domains, demonstrating superior performance and robustness compared to state-of-the-art techniques. These results underscore the potential of KANs in enhancing the reliability of machine learning systems in diverse environments.

Problem

• Detects out-of-distribution samples effectively
• Leverages local neuroplasticity of KANs
• Enhances machine learning system reliability

Innovation

• Local neuroplasticity in KANs
• Activation pattern comparison
• Superior OOD detection performance