mlsanityData to 14 Oct 2026

Large Language Models Are Zero-Shot Time Series Forecasters

arXiv 2310.07820 ↗First posted 11 Oct 2023cs.LGcited by 500

Not on the map. Why?

Cited by 31 in LLM reasoning, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

By encoding time series as a string of numerical digits, we can frame time series forecasting as next-token prediction in text. Developing this approach, we find that large language models (LLMs) such as GPT-3 and LLaMA-2 can surprisingly zero-shot extrapolate time series at a level comparable to or exceeding the performance of purpose-built time series models trained on the downstream tasks. To facilitate this performance, we propose procedures for effectively tokenizing time series data and converting discrete distributions over tokens into highly flexible densities over continuous values. We argue the success of LLMs for time series stems from their ability to naturally represent multimodal distributions, in conjunction with biases for simplicity, and repetition, which align with the salient features in many time series, such as repeated seasonal trends. We also show how LLMs can naturally handle missing data without imputation through non-numerical text, accommodate textual side information, and answer questions to help explain predictions. While we find that increasing model size generally improves performance on time series, we show GPT-4 can perform worse than GPT-3 because of how it tokenizes numbers, and poor uncertainty calibration, which is likely the result of alignment interventions such as RLHF.

Key facts. Cited by 500 papers in the map (data through 14 Oct 2026). Builds on 42 of its 91 references that are in the map. First posted on arXiv 11 Oct 2023.

Builds on

42 of 91 references.

References · 42

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2023GPT-4 Technical Reportcited by 18,396
  3. 2022Training language models to follow instructions with human feedbackcited by 15,971
  4. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 14,132
  5. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 13,028
  6. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  7. 2020Measuring Massive Multitask Language Understandingcited by 7,963
  8. 2017Deep reinforcement learning from human preferencescited by 3,912
  9. 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,780
  10. 2019The Curious Case of Neural Text Degenerationcited by 2,916
  11. 2023Sparks of Artificial General Intelligence: Early experiments with GPT-4cited by 2,592
  12. 2020Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecastingcited by 2,169
30 more
  1. 2016WaveNet: A Generative Model for Raw Audiocited by 1,722
  2. 2021Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecastingcited by 1,676
  3. 2022Are Transformers Effective for Time Series Forecasting?cited by 1,424
  4. 2022FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecastingcited by 1,210
  5. 2023PaLM 2 Technical Reportcited by 1,162
  6. 2017DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networkscited by 1,013
  7. 2022In-context Learning and Induction Headscited by 964
  8. 2021Show Your Work: Scratchpads for Intermediate Computation with Language Modelscited by 886
  9. 2019N-BEATS: Neural basis expansion analysis for interpretable time series forecastingcited by 753
  10. 2021Deduplicating Training Data Makes Language Models Bettercited by 714
  11. 2023One Fits All:Power General Time Series Analysis by Pretrained LMcited by 632
  12. 2018GPyTorch: Blackbox Matrix-Matrix Gaussian Process Inference with GPU Accelerationcited by 609
  13. 2023Language Modeling Is Compressioncited by 225
  14. 2016Temporal Convolutional Networks: A Unified Approach to Action Segmentationcited by 199
  15. 2022PromptCast: A New Prompt-based Learning Paradigm for Time Series Forecastingcited by 172
  16. 2023Emergent and Predictable Memorization in Large Language Modelscited by 148
  17. 2022N-HiTS: Neural Hierarchical Interpolation for Time Series Forecastingcited by 146
  18. 2023Meta-Transformer: A Unified Framework for Multimodal Learningcited by 136
  19. 2013Gaussian Process Kernels for Pattern Discovery and Extrapolationcited by 132
  20. 2023How well do Large Language Models perform in Arithmetic tasks?cited by 130
  21. 2023Speak, Memory: An Archaeology of Books Known to ChatGPT/GPT-4cited by 125
  22. 2021Can You Learn an Algorithm? Generalizing from Easy to Hard Problems with Recurrent Networkscited by 117
  23. 2021Darts: User-Friendly Modern Machine Learning for Time Seriescited by 91
  24. 2023CoLT5: Faster Long-Range Transformers with Conditional Computationcited by 84
  25. 2023Goat: Fine-tuned LLaMA Outperforms GPT-4 on Arithmetic Taskscited by 84
  26. 2023The No Free Lunch Theorem, Kolmogorov Complexity, and the Role of Inductive Biases in Machine Learningcited by 63
  27. 2022Forecast Evaluation for Data Scientists: Common Pitfalls and Best Practicescited by 55
  28. 2022The Lie Derivative for Measuring Learned Equivariancecited by 54
  29. 2022Path Independent Equilibrium Models Can Better Exploit Test-Time Computationcited by 38
  30. 2022Preformer: Predictive Transformer with Multi-Scale Segment-wise Correlations for Long-Term Time Series Forecastingcited by 23

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 10; 2024: 182; 2025: 170; 2026: 138.

500 papers.

Sort:

By area:

  1. 2024Chronos: Learning the Language of Time Seriescited by 759
  2. 2023A decoder-only foundation model for time-series forecastingcited by 692
  3. 2024Unified Training of Universal Time Series Forecasting Transformerscited by 614
  4. 2024MOMENT: A Family of Open Time-series Foundation Modelscited by 482
  5. 2024Foundation Models for Time Series Analysis: A Tutorial and Surveycited by 239
  6. 2024Deep Time Series Models: A Comprehensive Survey and Benchmarkcited by 197
  7. 2025Chronos-2: From Univariate to Universal Forecastingcited by 194
  8. 2025Sundial: A Family of Highly Capable Time Series Foundation Modelscited by 180
  9. 2024Are Language Models Actually Useful for Time Series Forecasting?cited by 168
  10. 2024Large Language Models(LLMs) on Tabular Data: Prediction, Generation, and Understanding -- A Surveycited by 159
  11. 2023UniTime: A Language-Empowered Unified Model for Cross-Domain Time Series Forecastingcited by 144
  12. 2024Tiny Time Mixers (TTMs): Fast Pre-trained Models for Enhanced Zero/Few-Shot Forecasting of Multivariate Time Seriescited by 127
38 more
  1. 2024TimeMixer++: A General Time Series Pattern Machine for Universal Predictive Analysiscited by 116
  2. 2024Large Language Models for Time Series: A Surveycited by 114
  3. 2023Large Models for Time Series and Spatio-Temporal Data: A Survey and Outlookcited by 110
  4. 2024ForecastBench: A Dynamic Benchmark of AI Forecasting Capabilitiescited by 110
  5. 2024Approaching Human-Level Forecasting with Language Modelscited by 107
  6. 2024CycleNet: Enhancing Time Series Forecasting through Modeling Periodic Patternscited by 106
  7. 2023ForecastPFN: Synthetically-Trained Zero-Shot Forecastingcited by 104
  8. 2024TimeCMA: Towards LLM-Empowered Multivariate Time Series Forecasting via Cross-Modality Alignmentcited by 103
  9. 2024From News to Forecast: Integrating Event Analysis in LLM-Based Time Series Forecasting with Reflectioncited by 98
  10. 2024Moirai-MoE: Empowering Time Series Foundation Models with Sparse Mixture of Expertscited by 95
  11. 2024UniTS: A Unified Multi-Task Time Series Modelcited by 94
  12. 2024Fine-Tuned Language Models Generate Stable Inorganic Materials as Textcited by 93
  13. 2024Language Models Still Struggle to Zero-shot Reason about Time Seriescited by 88
  14. 2024VisionTS: Visual Masked Autoencoders Are Free-Lunch Zero-Shot Time Series Forecasterscited by 88
  15. 2024Position: Graph Foundation Models are Already Herecited by 86
  16. 2024Spatial-Temporal Large Language Model for Traffic Predictioncited by 83
  17. 2024Large Language Model (LLM) for Telecommunications: A Comprehensive Survey on Principles, Key Techniques, and Opportunitiescited by 83
  18. 2024Large Scale Transfer Learning for Tabular Data via Language Modelingcited by 76
  19. 2024Time-MMD: Multi-Domain Multimodal Dataset for Time Series Analysiscited by 74
  20. 2025Test-time regression: a unifying framework for designing sequence models with associative memorycited by 74
  21. 2024Kolmogorov-Arnold Networks for Time Series: Bridging Predictive Power and Interpretabilitycited by 70
  22. 2024PaPaGei: Open Foundation Models for Optical Physiological Signalscited by 69
  23. 2025This Time is Different: An Observability Perspective on Time Series Foundation Modelscited by 68
  24. 2024Wisdom of the Silicon Crowd: LLM Ensemble Prediction Capabilities Rival Human Crowd Accuracycited by 67
  25. 2024AutoTimes: Autoregressive Time Series Forecasters via Large Language Modelscited by 65
  26. 2025Time-VLM: Exploring Multimodal Vision-Language Models for Augmented Time Series Forecastingcited by 63
  27. 2024Position: Bayesian Deep Learning is Needed in the Age of Large-Scale AIcited by 62
  28. 2024Empowering Time Series Analysis with Large Language Models: A Surveycited by 60
  29. 2024LSTPrompt: Large Language Models as Zero-Shot Time Series Forecasters by Long-Short-Term Promptingcited by 59
  30. 2024Is In-Context Learning in Large Language Models Bayesian? A Martingale Perspectivecited by 57
  31. 2025Deep Learning is Not So Mysterious or Differentcited by 56
  32. 2024Large Language Models for Forecasting and Anomaly Detection: A Systematic Literature Reviewcited by 53
  33. 2024FlowLLM: Flow Matching for Material Generation with Large Language Models as Base Distributionscited by 53
  34. 2024EEGFormer: Towards Transferable and Interpretable Large-Scale EEG Foundation Modelcited by 52
  35. 2024TimeSeriesExam: A time series understanding examcited by 49
  36. 2024A Sober Look at LLMs for Material Discovery: Are They Actually Good for Bayesian Optimization Over Molecules?cited by 48
  37. 2024Time Series Forecasting with LLMs: Understanding and Enhancing Model Capabilitiescited by 48
  38. 2024Toto: Time Series Optimized Transformer for Observabilitycited by 48

50 most cited of 500 All 500 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2023Time-LLM: Time Series Forecasting by Reprogramming Large Language Models
  2. 2026Rethinking the Role of LLMs in Time Series Forecasting
  3. 2025The language of time: a language model perspective on time-series foundation models
  4. 2025STELLA: Guiding Large Language Models for Time Series Forecasting with Semantic Abstractions
  5. 2024Understanding Why Large Language Models Can Be Ineffective in Time Series Analysis: The Impact of Modality Alignment
  6. 2025Conversational Time Series Foundation Models: Towards Explainable and Effective Forecasting
  7. 2025Small Vocabularies, Big Gains: Pretraining and Tokenization in Time Series Models
  8. 2025SciTS: Scientific Time Series Understanding and Generation with LLMs