AI RESEARCH
Non-Stationarity in the Embedding Space of Time Series Foundation Models
arXiv CS.LG
•
ArXi:2604.16428v1 Announce Type: new Time series foundation models (TSFMs) are widely used as generic feature extractors, yet the notion of non-stationarity in their embedding spaces remains poorly understood. Recent work often conflates non-stationarity with distribution shift, blurring distinctions fundamental to classical time-series analysis and long-standing methodologies such as statistical process control