Evaluating Uncertainty Calibration in Probabilistic Time Series Foundation Models
arXiv cs.AI · 3 Nov 2026, 04:34 AEST
Time series foundation models (TSFMs) are increasingly used for probabilistic forecasting, but their uncertainty estimates are not well understood. This work introduces a framework for evaluating uncertainty calibration in TSFMs and reveals systematic patterns of miscalibration in current models.
Key facts
- In this work, we study uncertainty calibration in probabilistic TSFMs from a broader methodological and empirical perspective.
- We introduce a taxonomy of how TSFMs represent predictive uncertainty and develop a fine-grained evaluation framework for calibration in multi-horizon forecasting.
- Using this framework, we conduct a comprehensive empirical evaluation of popular TSFMs across diverse datasets and show that current models exhibit distinct and systematic patterns of miscalibration.
Read aloud
Time series foundation models TSFMs are increasingly used for probabilistic forecasting, but their uncertainty estimates are not well understood. This work introduces a framework for evaluating uncertainty calibration in TSFMs and reveals systematic patterns of miscalibration in current models. In this work, we study uncertainty calibration in probabilistic TSFMs from a broader methodological and empirical perspective. We introduce a taxonomy of how TSFMs represent predictive uncertainty and develop a fine-grained evaluation framework for calibration in multi-horizon forecasting. Using this framework, we conduct a comprehensive empirical evaluation of popular TSFMs across diverse datasets and show that current models exhibit distinct and systematic patterns of miscalibration.
Sources in this cluster
- Evaluating Uncertainty Calibration in Probabilistic Time Series Foundation Models
arXiv cs.AI · 3 Nov 2026, 04:34 AEST