Interpretability and Root-Cause Analysis for Time Series

"Being right was not enough; they needed me to say, in plain terms, which wiggle in the signal made up my mind."

An Explainable Embodied AI Agent

Placeholder chapter page. Sections are produced by the book-skills pipeline.

Sections

1. 67.1 Attribution for time series (Integrated Gradients, SHAP-TS)
2. 67.2 Temporal saliency and its pitfalls
3. 67.3 Counterfactual sensor explanations (TSEvo, MASCOTS)
4. 67.4 Concept- and prototype-based explanations
5. 67.5 Root-cause graphs
6. 67.6 Human-review tooling (TSInterpret, Captum)
7. 67.7 Explanation faithfulness and benchmarks (XTSC-Bench)