# DiFA: Dual Evidence Fusion and Aggregation for Token-Level Text Anomaly Detection

arXiv: https://arxiv.org/abs/2609.22136 · HTML: https://mlsanity.com/paper/2609.22136/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 26 of its 41 references that are in the map.
- First posted on arXiv 24 Aug 2026.

## Builds on (most cited first, 26 of 26)

- [ECOD: Unsupervised Outlier Detection Using Empirical Cumulative Distribution Functions](https://mlsanity.com/paper/2201.00382/) (2022)
- [LUNAR: Unifying Local Outlier Detection Methods via Graph Neural Networks](https://mlsanity.com/paper/2112.05355/) (2021)
- [Deep Graph Anomaly Detection: A Survey and New Perspectives](https://mlsanity.com/paper/2409.09957/) (2024)
- [DATE: Detecting Anomalies in Text via Self-Supervision of Transformers](https://mlsanity.com/paper/2104.05591/) (2021)
- [Adaptive pooling operators for weakly labeled sound event detection](https://mlsanity.com/paper/1804.10070/) (2018)
- [Generative Semi-supervised Graph Anomaly Detection](https://mlsanity.com/paper/2402.11887/) (2024)
- [NLP-ADBench: NLP Anomaly Detection Benchmark](https://mlsanity.com/paper/2412.04784/) (2024)
- [DynHD: Hallucination Detection for Diffusion Large Language Models via Denoising Dynamics Deviation Learning](https://mlsanity.com/paper/2603.16459/) (2026)
- [OFA-MAS: One-for-All Multi-Agent System Topology Design based on Mixture-of-Experts Graph Generative Models](https://mlsanity.com/paper/2601.12996/) (2026)
- [Few-shot Anomaly Detection in Text with Deviation Learning](https://mlsanity.com/paper/2308.11780/) (2023)
- [Anomaly Detection Based on Isolation Mechanisms: A Survey](https://mlsanity.com/paper/2403.10802/) (2024)
- [Raising the Bar in Graph OOD Generalization: Invariant Learning Beyond Explicit Environment Modeling](https://mlsanity.com/paper/2502.10706/) (2025)
- [From Few-Shot to Zero-Shot: Towards Generalist Graph Anomaly Detection](https://mlsanity.com/paper/2602.18793/) (2026)
- [Influence-Oriented Personalized Federated Learning](https://mlsanity.com/paper/2410.03315/) (2024)
- [TAD-Bench: A Comprehensive Benchmark for Embedding-Based Text Anomaly Detection](https://mlsanity.com/paper/2501.11960/) (2025)
- [FedCIGAR: A Personalized Reconstruction Approach for Federated Graph-level Anomaly Detection](https://mlsanity.com/paper/2605.09428/) (2026)
- [Multi-Stage Verification-Centric Framework for Mitigating Hallucination in Multi-Modal RAG](https://mlsanity.com/paper/2507.20136/) (2025)
- [Text Anomaly Detection with Simplified Isolation Kernel](https://mlsanity.com/paper/2510.13197/) (2025)
- [Towards Anomaly Detection on Relational Data](https://mlsanity.com/paper/2606.18621/) (2026)
- [Rethinking Feature Alignment in Generalist Graph Anomaly Detection: A Relational Fingerprint-based Approach](https://mlsanity.com/paper/2605.25429/) (2026)
- [Semantics or spelling? Probing contextual word embeddings with orthographic noise](https://mlsanity.com/paper/2408.04162/) (2024)
- [Beyond a Single Perspective: Text Anomaly Detection with Multi-View Language Representations](https://mlsanity.com/paper/2601.17786/) (2026)
- [CAMERA: Adapting to Semantic Camouflage in Unsupervised Text-Attributed Graph Fraud Detection](https://mlsanity.com/paper/2605.20032/) (2026)
- [Advancing Pre-trained Teacher: Towards Robust Feature Discrepancy for Anomaly Detection](https://mlsanity.com/paper/2405.02068/) (2024)
- [Towards Token-Level Text Anomaly Detection](https://mlsanity.com/paper/2601.13644/) (2026)
- [Explained anomaly detection in text reviews: Can subjective scenarios be correctly evaluated?](https://mlsanity.com/paper/2311.04948/) (2023)

## Abstract (arXiv)

Text anomaly detection, the task of identifying text instances that deviate from normal language patterns, is crucial for language-driven applications. However, most existing methods can only perform document-level anomaly detection, making it hard to locate harmful phrases or support targeted prevention. Recently, there has been an emerging trend toward token-level text anomaly detection, which aims to address the above limitation by identifying anomalous words or fragments within a document. Nevertheless, one representative method mainly relies on representation-space distance measurement, neglecting the complementary roles of different anomaly cues in capturing diverse abnormal patterns. To bridge the gaps, we propose a Dual-evidence framework with adaptive Fusion and Aggregation (DiFA) for token-level anomaly detection. DiFA derives anomaly scores from form-structural and semantic views to capture visible structural abnormality and contextual inconsistency, respectively, thereby providing complementary evidence for identifying diverse anomalies. To combine these two scores with varying numerical scales, DiFA incorporates a calibration and fusion mechanism to adaptively balance the two views. Moreover, to obtain a discriminative document-level score, a multivariate aggregation method is designed to summarize token-level anomaly scores from multiple perspectives, preventing rare anomalous tokens from being diluted. Extensive experiments across various text anomaly detection benchmarks demonstrate that DiFA consistently achieves top performance while maintaining strong efficiency, robustness, and interpretability. The code and scripts are available at: https://github.com/qyy11-com/DiFA.