Halt the Hallucination: Decoupling Signal and Semantic OOD Detection Based on Cascaded Early Rejection

2026-02-06

AI Summary

This work addresses the inefficiency and semantic hallucination issues in existing out-of-distribution (OOD) detection methods, which perform full inference even on low-level noisy inputs. To overcome this, the authors propose a Cascaded Early Rejection (CER) framework that decouples signal-level and semantic-level anomaly detection for the first time. The framework integrates a Structural Energy Sieve (SES), based on the Laplacian operator, at the input stage and a hyperspherical energy detector (SHE) at intermediate layers. Evaluated on CIFAR-100, CER reduces the false positive rate at 95% true positive rate (FPR95) from 33.58% to 22.84%, achieves an AUROC of 93.97%, and cuts computational overhead by 32%. It demonstrates significant performance gains over current methods in real-world scenarios such as sensor failure detection.

Technology Category

• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Computer Vision: Object Detection & Categorization
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments

Abstract

Efficient and robust Out-of-Distribution (OOD) detection is paramount for safety-critical applications.However, existing methods still execute full-scale inference on low-level statistical noise. This computational mismatch not only incurs resource waste but also induces semantic hallucination, where deep networks forcefully interpret physical anomalies as high-confidence semantic features.To address this, we propose the Cascaded Early Rejection (CER) framework, which realizes hierarchical filtering for anomaly detection via a coarse-to-fine logic.CER comprises two core modules: 1)Structural Energy Sieve (SES), which establishes a non-parametric barrier at the network entry using the Laplacian operator to efficiently intercept physical signal anomalies; and 2) the Semantically-aware Hyperspherical Energy (SHE) detector, which decouples feature magnitude from direction in intermediate layers to identify fine-grained semantic deviations. Experimental results demonstrate that CER not only reduces computational overhead by 32% but also achieves a significant performance leap on the CIFAR-100 benchmark:the average FPR95 drastically decreases from 33.58% to 22.84%, and AUROC improves to 93.97%. Crucially, in real-world scenarios simulating sensor failures, CER exhibits performance far exceeding state-of-the-art methods. As a universal plugin, CER can be seamlessly integrated into various SOTA models to provide performance gains.

Problem

• Out-of-Distribution detection
• semantic hallucination
• computational inefficiency
• signal anomaly
• semantic deviation

Innovation

• Cascaded Early Rejection
• Out-of-Distribution Detection
• Semantic Hallucination
• Structural Energy Sieve
• Hyperspherical Energy