Review for NeurIPS paper: Benchmarking Deep Learning Interpretability in Time Series Predictions

This method is not developed for time-series, only for tabular data, so it is unclear how they would do this evaluation, neither is it described anywhere in the main draft.