Interventional Time Series Priors for Causal Foundation Models

Dennis Thumm, Ying Chen

Prior-data fitted networks (PFNs) have emerged as powerful foundation models for tabular causal inference, yet their extension to time series remains limited by the absence of synt…