mlsanityData to 14 Oct 2026

Semantically-Guided Inference for Conditional Diffusion Models: Enhancing Covariate Consistency in Time Series Forecasting

arXiv 2508.01761 ↗First posted 3 Aug 2025cs.LG cs.AIcited by 0

On the map: Diffusion models → Flow matching, time series and science

Abstract (arXiv)

Diffusion models have demonstrated strong performance in time series forecasting, yet often suffer from semantic misalignment between generated trajectories and conditioning covariates, especially under complex or multimodal conditions. To address this issue, we propose SemGuide, a plug-and-play, inference-time method that enhances covariate consistency in conditional diffusion models. Our approach introduces a scoring network to assess the semantic alignment between intermediate diffusion states and future covariates. These scores serve as proxy likelihoods in a stepwise importance reweighting procedure, which progressively adjusts the sampling path without altering the original training process. The method is model-agnostic and compatible with any conditional diffusion framework. Experiments on real-world forecasting tasks show consistent gains in both predictive accuracy and covariate alignment, with especially strong performance under complex conditioning scenarios.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Flow matching, time series and science, one of 14 directions in Diffusion models. Builds on 15 of its 24 references that are in the map. First posted on arXiv 3 Aug 2025.

Builds on

15 of 24 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024MTSCI: A Conditional Diffusion Model for Multivariate Time Series Consistent Imputation11 Aug
  2. 2024Diffusion-TS: Interpretable Diffusion for General Time Series Generation4 Mar
  3. 2021CSDI: Conditional Score-based Diffusion Models for Probabilistic Time Series Imputation7 Jul
  4. 2021Autoregressive Denoising Diffusion Models for Multivariate Probabilistic Time Series Forecasting28 Jan

Other references · 11

  1. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  3. 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,255
  4. 2022Classifier-Free Diffusion Guidancecited by 5,847
  5. 2022Diffusion Models in Vision: A Surveycited by 1,015
  6. 2023iTransformer: Inverted Transformers Are Effective for Time Series Forecastingcited by 931
  7. 2019Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecastingcited by 761
  8. 2023Long-term Forecasting with TiDE: Time-series Dense Encodercited by 332
  9. 2024TimeXer: Empowering Transformers for Time Series Forecasting with Exogenous Variablescited by 248
  10. 2020Forecasting day-ahead electricity prices: A review of state-of-the-art algorithms, best practices and an open-access benchmarkcited by 108
  11. 2021Time Series Prediction using Deep Learning Methods in Healthcarecited by 74

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2023Predict, Refine, Synthesize: Self-Guiding Diffusion Models for Probabilistic Time Series Forecasting
  2. 2025Conditionally Whitened Generative Models for Probabilistic Time Series Forecasting
  3. 2025Diffusion Models for Time Series Forecasting: A Survey
  4. 2026When Denoising Hurts: Rethinking the Terminal Step of Diffusion Time Series Forecasters -- Extended Version
  5. 2025UniDiff: A Unified Diffusion Framework for Multimodal Time Series Forecasting
  6. 2026GARDiff: Graph-Aligned Residual Diffusion for Probabilistic Multivariate Time-Series Forecasting
  7. 2026Differencing the Diffusion Trajectory toward Uncertain Components for Time Series Forecasting
  8. 2024The Rise of Diffusion Models in Time-Series Forecasting