![[2502.17812] Can Multimodal LLMs Perform Time Series Anomaly Detection?](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2502.17812] Can Multimodal LLMs Perform Time Series Anomaly Detection?
Summary
The paper explores the potential of multimodal large language models (MLLMs) for time series anomaly detection (TSAD), introducing a new benchmark and framework for automatic detection.
Why It Matters
Time series anomaly detection is crucial for maintaining the reliability of online systems. This research highlights the untapped potential of MLLMs in enhancing TSAD, which could lead to more effective monitoring and fault diagnosis in various applications.
Key Takeaways
- MLLMs can enhance time series anomaly detection beyond traditional methods.
- The VisualTimeAnomaly benchmark provides a comprehensive evaluation of MLLMs for TSAD.
- The proposed TSAD-Agents framework automates detection through collaborative agents.
- The study addresses limitations in existing TSAD approaches by considering multi-granular and irregular anomalies.
- Results indicate significant improvements in detection performance using multimodal approaches.
Computer Science > Computation and Language arXiv:2502.17812 (cs) [Submitted on 25 Feb 2025 (v1), last revised 17 Feb 2026 (this version, v2)] Title:Can Multimodal LLMs Perform Time Series Anomaly Detection? Authors:Xiongxiao Xu, Haoran Wang, Yueqing Liang, Philip S. Yu, Yue Zhao, Kai Shu View a PDF of the paper titled Can Multimodal LLMs Perform Time Series Anomaly Detection?, by Xiongxiao Xu and 5 other authors View PDF HTML (experimental) Abstract:Time series anomaly detection (TSAD) has been a long-standing pillar problem in Web-scale systems and online infrastructures, such as service reliability monitoring, system fault diagnosis, and performance optimization. Large language models (LLMs) have demonstrated unprecedented capabilities in time series analysis, the potential of multimodal LLMs (MLLMs), particularly vision-language models, in TSAD remains largely under-explored. One natural way for humans to detect time series anomalies is through visualization and textual description. It motivates our research question: Can multimodal LLMs perform time series anomaly detection? Existing studies often oversimplify the problem by treating point-wise anomalies as special cases of range-wise ones or by aggregating point anomalies to approximate range-wise scenarios. They limit our understanding for realistic scenarios such as multi-granular anomalies and irregular time series. To address the gap, we build a VisualTimeAnomaly benchmark to comprehensively investigate zero-shot...