AI RESEARCH
Causal-INSIGHT: Probing Temporal Models to Extract Causal Structure
arXiv CS.LG
•
ArXi:2603.25473v1 Announce Type: new Understanding directed temporal interactions in multivariate time series is essential for interpreting complex dynamical systems and the predictive models trained on them. We present Causal-INSIGHT, a model-agnostic, post-hoc interpretation framework for extracting model-implied (predictor-dependent), directed, time-lagged influence structure from trained temporal predictors.