The Failures of Marginal Influence-Based Attribution Methods for Global Time Series Explanations

2026/06/25 by Amadeo Tunyi
#cs.LG #cs.AI

paper · pdf

Abstract

Explainability methods for time series models predominantly produce flat attribution scores: they quantify the direct influence of a feature at a timestamp by a scalar. We prove that the dominant failure mode of such methods is not the scalar format itself but a fundamental computational mismatch: existing methods compute scores via marginal conditioning or off-manifold gradients, both of which conflate direct temporal dependencies with mediated ones under autocorrelation. We also define DAG-faithfulness: an explanation is DAG-faithful if the temporal dependency graph it encodes is Markov-equivalent to the temporal directed acyclic graph (DAG) implicitly learned by the model. Particularly, we observe that standard attribution methods, specifically SHAP, are not DAG-faithful in general, and that recent time-series-aware extensions inherit the same computational limitation.

Citations

• Attention Is All You Need

Related

• TIMING: Temporality-Aware Integrated Gradients for Time Series Explanation
• Explaining Time Series Classification Predictions via Causal Attributions
• TSInsight: A local-global attribution framework for interpretability in time-series data
• An Empirical Evaluation of Factors Affecting SHAP Explanation of Time Series Classification
• Time Series Model Attribution Visualizations as Explanations