Learning from Black-Box Models by Maximum Interpretation Decomposition R package midr version 0.6.1 The goal of 'midr' is to provide a model-agnostic method for interpreting and explaining black-box predictive models by creating a globally interpretable surrogate model. The package implements 'Maximum Interpretation Decomposition' (MID), a functional decomposition technique that finds an optimal additive approximation of the original model. This approximation is achieved by minimizing the squared error between the predictions of the black-box model and the surrogate model. The theoretical foundations of MID are described in Iwasawa & Matsumori (2025) [Forthcoming], and the package itself is detailed in Asashiba et al. (2025) <doi:10.48550/arXiv.2506.08338>. Software Imports: graphics, grDevices, Rcpp, RcppEigen, rlang, stats, utils LinkingTo: Rcpp Suggests: datasets, ggplot2, khroma, knitr, RColorBrewer, rmarkdown, scales, testthat, viridisLite Ryoichi Asashiba <ryoichi.asashiba@gmail.com> Comprehensive R Archive Network (CRAN) Ryoichi Asashiba [aut, cre] (ORCID: <https://orcid.org/0009-0001-9532-7000>), Hirokazu Iwasawa [aut], Reiji Kozuma [ctb] MIT + file LICENSE (https://CRAN.R-project.org/package=midr/LICENSE) 2026-05-01 application/tgz https://CRAN.R-project.org/package=midr doi:10.32614/CRAN.package.midr