English

AD-LLM: Benchmarking Large Language Models for Anomaly Detection

Computation and Language 2025-10-13 v4 Artificial Intelligence

Abstract

Anomaly detection (AD) is an important machine learning task with many real-world uses, including fraud detection, medical diagnosis, and industrial monitoring. Within natural language processing (NLP), AD helps detect issues like spam, misinformation, and unusual user activity. Although large language models (LLMs) have had a strong impact on tasks such as text generation and summarization, their potential in AD has not been studied enough. This paper introduces AD-LLM, the first benchmark that evaluates how LLMs can help with NLP anomaly detection. We examine three key tasks: (i) zero-shot detection, using LLMs' pre-trained knowledge to perform AD without tasks-specific training; (ii) data augmentation, generating synthetic data and category descriptions to improve AD models; and (iii) model selection, using LLMs to suggest unsupervised AD models. Through experiments with different datasets, we find that LLMs can work well in zero-shot AD, that carefully designed augmentation methods are useful, and that explaining model selection for specific datasets remains challenging. Based on these results, we outline six future research directions on LLMs for AD.

Keywords

Cite

@article{arxiv.2412.11142,
  title  = {AD-LLM: Benchmarking Large Language Models for Anomaly Detection},
  author = {Tiankai Yang and Yi Nian and Shawn Li and Ruiyao Xu and Yuangang Li and Jiaqi Li and Zhuo Xiao and Xiyang Hu and Ryan Rossi and Kaize Ding and Xia Hu and Yue Zhao},
  journal= {arXiv preprint arXiv:2412.11142},
  year   = {2025}
}

Comments

Accepted to Findings of the Association for Computational Linguistics (ACL 2025), Vienna, Austria

R2 v1 2026-06-28T20:35:45.191Z