mlsanityData to 7 Oct 2026

Can Multimodal LLMs Perform Time Series Anomaly Detection?

arXiv 2502.17812 ↗First posted 25 Feb 2025cs.CL cs.LGcited by 11

On the map: Multimodal LLMs / vision-language models → Chart understanding and anomaly detection

Abstract (arXiv)

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 capabilities of MLLMs for TSAD, progressively from point-, range-, to variate-wise anomalies, and extends to irregular sampling conditions. Our study reveals several key insights in multimodal MLLMs for TSAD. Built on these findings, we propose a MLLMs-based multi-agent framework TSAD-Agents to achieve automatic TSAD. Our framework comprises scanning, planning, detection, and checking agents that synergistically collaborate to reason, plan, and self-reflect to enable automatic TSAD. These agents adaptively invoke tools such as traditional methods and MLLMs and dynamically switch between text and image modalities to optimize detection performance.

Key facts. Cited by 11 papers in the map (data through 7 Oct 2026). Placed in Chart understanding and anomaly detection, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 47 of its 87 references that are in the map. First posted on arXiv 25 Feb 2025.

Builds on

47 of 87 references.

References · 47

  1. 2023GPT-4 Technical Reportcited by 18,132
  2. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,082
  3. 2023Visual Instruction Tuningcited by 7,792
  4. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolutioncited by 4,151
  5. 2024Gemini 1.5: Unlocking multimodal understanding across millions of tokens of contextcited by 3,175
  6. 2024Video-MME: The First-Ever Comprehensive Evaluation Benchmark of Multi-modal LLMs in Video Analysiscited by 1,375
  7. 2023SEED-Bench: Benchmarking Multimodal LLMs with Generative Comprehensioncited by 929
  8. 2023A Survey on Multimodal Large Language Modelscited by 898
  9. 2023One Fits All:Power General Time Series Analysis by Pretrained LMcited by 615
  10. 2024Unified Training of Universal Time Series Forecasting Transformerscited by 588
  11. 2023SpeechGPT: Empowering Large Language Models with Intrinsic Cross-Modal Conversational Abilitiescited by 518
  12. 2018The UCR Time Series Archivecited by 421
35 more
  1. 2018The UEA multivariate time series classification archive, 2018cited by 330
  2. 2024Long-context LLMs Struggle with Long In-context Learningcited by 285
  3. 2023Challenges and Applications of Large Language Modelscited by 251
  4. 2024MA-LMM: Memory-Augmented Large Multimodal Model for Long-Term Video Understandingcited by 223
  5. 2023Pengi: An Audio Language Model for Audio Taskscited by 217
  6. 2023OpenAGI: When LLM Meets Domain Expertscited by 196
  7. 2022Deep Learning for Time Series Anomaly Detection: A Surveycited by 130
  8. 2023BLIVA: A Simple Multimodal LLM for Better Handling of Text-Rich Visual Questionscited by 125
  9. 2023Table Meets LLM: Can Large Language Models Understand Structured Table Data? A Benchmark and Empirical Studycited by 116
  10. 2024ChatTime: A Unified Multimodal Time Series Foundation Model Bridging Numerical and Textual Datacited by 92
  11. 2024VisionTS: Visual Masked Autoencoders Are Free-Lunch Zero-Shot Time Series Forecasterscited by 87
  12. 2022Local Evaluation of Time Series Anomaly Detection Algorithmscited by 79
  13. 2023Time Series as Images: Vision Transformer for Irregularly Sampled Time Seriescited by 79
  14. 2024LSTPrompt: Large Language Models as Zero-Shot Time Series Forecasters by Long-Short-Term Promptingcited by 56
  15. 2024A Survey on Evaluation of Multimodal Large Language Modelscited by 46
  16. 2020A Survey on Principles, Models and Methods for Learning from Irregularly Sampled Time Seriescited by 42
  17. 2023ChatGPT Informed Graph Neural Network for Stock Movement Predictioncited by 41
  18. 2024Can LLMs Understand Time Series Anomalies?cited by 41
  19. 2024Number Cookbook: Number Understanding of Language Models and How to Improve Itcited by 41
  20. 2023GATGPT: A Pre-trained Large Language Model with Graph Attention Network for Spatiotemporal Imputationcited by 35
  21. 2024TSI-Bench: Benchmarking Time Series Imputationcited by 33
  22. 2024Large language models can be zero-shot anomaly detectors for time series?cited by 30
  23. 2024Can Knowledge Editing Really Correct Hallucinations?cited by 28
  24. 2024Large Language Model Guided Knowledge Distillation for Time Series Anomaly Detectioncited by 26
  25. 2024Large Language Models can Deliver Accurate and Interpretable Time Series Anomaly Detectioncited by 26
  26. 2024AD-LLM: Benchmarking Large Language Models for Anomaly Detectioncited by 25
  27. 2024Can LLMs Serve As Time Series Anomaly Detectors?cited by 22
  28. 2024See it, Think it, Sorted: Large Multimodal Models are Few-shot Time Series Anomaly Analyzerscited by 20
  29. 2022Constructing Large-Scale Real-World Benchmark Datasets for AIOpscited by 19
  30. 2024Graph Mixture of Experts and Memory-augmented Routers for Multivariate Time Series Anomaly Detectioncited by 19
  31. 2023Weather2K: A Multivariate Spatio-Temporal Benchmark Dataset for Meteorological Forecasting Based on Real-Time Observation Data from Ground Weather Stationscited by 18
  32. 2024Multimodal Large Language Model is a Human-Aligned Annotator for Text-to-Image Generationcited by 15
  33. 2024Taxonomy-Guided Zero-Shot Recommendations with LLMscited by 15
  34. 2024Piecing It All Together: Verifying Multi-Hop Multimodal Claimscited by 12
  35. 2024Plots Unlock Time-Series Understanding in Multimodal Modelscited by 9

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 7; 2026: 4.

11 papers.

  1. 2025A Survey of Reasoning and Agentic Systems in Time Series with Large Language Modelscited by 21
  2. 2025JailDAM: Jailbreak Detection with Adaptive Memory for Vision-Language Modelcited by 10
  3. 2025Harnessing Vision-Language Models for Time Series Anomaly Detectioncited by 9
  4. 2025Agentic AI for autonomous anomaly management in complex systemscited by 4
  5. 2025CALM: A Framework for Continuous, Adaptive, and LLM-Mediated Anomaly Detection in Time-Series Streamscited by 4
  6. 2026Detecting Time Series Anomalies Like an Expert: A Multi-Agent LLM Framework with Specialized Analyzerscited by 1
  7. 2025Evaluating Simplification Algorithms for Interpretability of Time Series Classificationcited by 0
  8. 2025ViTs: Teaching Machines to See Time Series Anomalies Like Human Expertscited by 0
  9. 2026From Consistency to Complementarity: Aligned and Disentangled Multi-modal Learning for Time Series Understanding and Reasoningcited by 0
  10. 2026AnomSeer: Reinforcing Multimodal LLMs to Reason for Time-Series Anomaly Detectioncited by 0
  11. 2026VETime: Vision Enhanced Zero-Shot Time Series Anomaly Detectioncited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Refining Time Series Anomaly Detectors using Large Language Models
  2. 2025AXIS: Explainable Time Series Anomaly Detection with Large Language Models
  3. 2026MMR-AD: A Large-Scale Multimodal Dataset for Benchmarking General Anomaly Detection with Multimodal Large Language Models
  4. 2026Evaluating Large Language Models for Time Series Anomaly Detection in Aerospace Software
  5. 2026ChatAD: Reasoning-Enhanced Time-Series Anomaly Detection with Multi-Turn Instruction Evolution
  6. 2026LEARN-TS: LLM-Enhanced Alignment and Reconstruction with Normality Guidance for Multivariate Time-Series Anomaly Detection
  7. 2024Quo Vadis, Anomaly Detection? LLMs and VLMs in the Spotlight
  8. 2026Are Multimodal LLMs Ready for Surveillance? A Reality Check on Zero-Shot Anomaly Detection in the Wild