[2603.25473] Causal-INSIGHT: Probing Temporal Models to Extract Causal Structure

Computer Science > Machine Learning arXiv:2603.25473 (cs) [Submitted on 26 Mar 2026] Title:Causal-INSIGHT: Probing Temporal Models to Extract Causal Structure Authors:Benjamin Redden, Hui Wang, Shuyan Li View a PDF of the paper titled Causal-INSIGHT: Probing Temporal Models to Extract Causal Structure, by Benjamin Redden and 2 other authors View PDF HTML (experimental) Abstract: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. Rather than inferring causal structure at the level of the data-generating process, Causal-INSIGHT analyzes how a fixed, pre-trained predictor responds to systematic, intervention-inspired input clamping applied at inference time. From these responses, we construct directed temporal influence signals that reflect the dependencies the predictor relies on for prediction, and introduce Qbic, a sparsity-aware graph selection criterion that balances predictive fidelity and structural complexity without requiring ground-truth graph labels. Experiments across synthetic, simulated, and realistic benchmarks show that Causal-INSIGHT generalizes across diverse backbone architectures, maintains competitive structural accuracy, and yields s...