AI RESEARCH
Towards Continuous-time Causal Foundation Models
arXiv CS.LG
•
ArXi:2605.28880v1 Announce Type: new Extending discrete-time causal Prior-data Fitted Networks for time series to continuous time invites writing the mechanism as a stochastic differential equation (SDE) -- but if the SDE is integrated \emph{once per observation gap}, the trajectory law depends on when it is observed, and the prior remains a discrete-time Marko model in SDE clothing.