mlsanityData to 7 Oct 2026

Distribution-Aware Diffusion-LLM for Robust Ultra-Long-Term Time Series Forecasting

arXiv 2606.23391 ↗First posted 22 Jun 2026cs.LG cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Time series forecasting is a fundamental machine learning task. Recent work has explored Large Language Models (LLMs) for this purpose due to their strong generalization, pattern recognition, and zero-shot or few-shot capabilities. Despite their suitability for long-context learning, LLMs face challenges in multimodal settings: they lack calibrated probabilistic modeling for non-text data and struggle to align heterogeneous representations. To address these issues, we propose a new framework Diffusion-LLM that integrates a conditional diffusion model into an LLM-based forecasting pipeline. This joint design enables learning the conditional distribution of future data while improving semantic alignment in a shared latent space. We evaluate Diffusion-LLM on six long-term forecasting benchmarks, including ETT, Weather, and ECL. Our method consistently outperforms existing LLM-based baseline, achieving notable gains in ultra-long-term and few-shot forecasting and demonstrating the value of distribution-aware regularization for enhancing robustness and generalization in time series LLMs.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 36 of its 46 references that are in the map. First posted on arXiv 22 Jun 2026.

Builds on

36 of 46 references.

References · 36

  1. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  2. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  3. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 13,958
  4. 2020Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecastingcited by 2,141
  5. 2022A Time Series is Worth 64 Words: Long-term Forecasting with Transformerscited by 1,981
  6. 2023ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIscited by 1,880
  7. 2020Reformer: The Efficient Transformercited by 1,658
  8. 2021Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecastingcited by 1,654
  9. 2023A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questionscited by 1,582
  10. 2022Are Transformers Effective for Time Series Forecasting?cited by 1,400
  11. 2022FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecastingcited by 1,200
  12. 2022TimesNet: Temporal 2D-Variation Modeling for General Time Series Analysiscited by 1,030
24 more
  1. 2023Time-LLM: Time Series Forecasting by Reprogramming Large Language Modelscited by 680
  2. 2023One Fits All:Power General Time Series Analysis by Pretrained LMcited by 615
  3. 2022Non-stationary Transformers: Exploring the Stationarity in Time Series Forecastingcited by 385
  4. 2021Autoregressive Denoising Diffusion Models for Multivariate Probabilistic Time Series Forecastingcited by 332
  5. 2023Trusting Your Evidence: Hallucinate Less with Context-aware Decodingcited by 319
  6. 2017A Multi-Horizon Quantile Recurrent Forecastercited by 211
  7. 2022PromptCast: A New Prompt-based Learning Paradigm for Time Series Forecastingcited by 170
  8. 2023TEST: Text Prototype Aligned Embedding to Activate LLM's Ability for Time Seriescited by 165
  9. 2022Less Is More: Fast Multivariate Time Series Forecasting with Light Sampling-oriented MLP Structurescited by 158
  10. 2022ETSformer: Exponential Smoothing Transformers for Time-series Forecastingcited by 146
  11. 2023LLM4TS: Aligning Pre-Trained LLMs as Data-Efficient Time-Series Forecasterscited by 127
  12. 2020Deep Learning for Time Series Forecasting: Tutorial and Literature Surveycited by 117
  13. 2023Temporal Data Meets LLM -- Explainable Financial Time Series Forecastingcited by 77
  14. 2021Time Series Prediction using Deep Learning Methods in Healthcarecited by 71
  15. 2025Time-VLM: Exploring Multimodal Vision-Language Models for Augmented Time Series Forecastingcited by 59
  16. 2021ScoreGrad: Multivariate Probabilistic Time Series Forecasting with Continuous Energy-based Generative Modelscited by 42
  17. 2024TOTEM: TOkenized Time Series EMbeddings for General Time Series Analysiscited by 37
  18. 2023ETP: Learning Transferable ECG Representations via ECG-Text Pre-trainingcited by 28
  19. 2023TENT: Connect Language Models with IoT Sensors for Zero-Shot Activity Recognitioncited by 25
  20. 2023Leveraging Vision-Language Models for Granular Market Change Predictioncited by 22
  21. 2024Implicit Multimodal Alignment: On the Generalization of Frozen LLMs to Multimodal Inputscited by 22
  22. 2025Vision-Enhanced Time Series Forecasting via Latent Diffusion Modelscited by 9
  23. 2023Trade-offs in Fine-tuned Diffusion Models Between Accuracy and Interpretabilitycited by 8
  24. 2023Predicting Battery Lifetime Under Varying Usage Conditions from Early Aging Datacited by 7

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Enhancing few-shot time series forecasting with LLM-guided diffusion
  2. 2026Rethinking the Role of LLMs in Time Series Forecasting
  3. 2025Fusing Large Language Models with Temporal Transformers for Time Series Forecasting
  4. 2025Adapting LLMs to Time Series Forecasting via Temporal Heterogeneity Modeling and Representation Alignment
  5. 2024NeST: Neighborhood-aware semantic alignment and temporal modulation for LLM based time series forecasting
  6. 2026Speaking Numbers to LLMs: Multi-Wavelet Number Embeddings for Time Series Forecasting
  7. 2025UniDiff: A Unified Diffusion Framework for Multimodal Time Series Forecasting
  8. 2024AutoTimes: Autoregressive Time Series Forecasters via Large Language Models