Credits go to
Mikolaj Konarski, the original author of the library
Conal Elliott, talking beautifully about the Essence of Automatic Differentiation
Simon Peyton Jones, who sketched in "Provably correct, asymptotically efficient, higher-order reverse-mode automatic differentiation" talk at https://www.youtube.com/watch?v=EPGqzkEZWyw algorithms that this Haskell codebase implements
Faustyna Krawiec, Neel Krishnaswami, Tom Ellis, Andrew Fitzgibbon, Richard Eisenberg, the remaining authors of the paper "Provably correct, asymptotically efficient, higher-order reverse-mode automatic differentiation" from POPL 2022 that describes and proves correct a comprehensive AD formalism, including some ingenious algorithmic ideas this Haskell codebase implements
Oleg Grenrus, whose https://hackage.haskell.org/package/overloaded contain inspiring examples related to AD
Edward Kmett, author of the archetype Haskell https://github.com/ekmett/ad library
Justin Le, author of the https://github.com/mstksg/backprop library and the amazing articles at https://blog.jle.im
Tom Smeding, who influenced the major implementation decisions of horde-ad and solved the hardest technical problems
and other benefactors and contributors, in chronological order: Stanisław Findeisen, Cale Gibbard, Well-Typed LLP
Authors of files t10k-images-idx3-ubyte.gz, t10k-labels-idx1-ubyte.gz, train-images-idx3-ubyte.gz, train-labels-idx1-ubyte.gz from http://yann.lecun.com/exdb/mnist, copied into this repo into directory samplesData/, copyright (according to https://keras.io/api/datasets/mnist) Yann LeCun and Corinna Cortes for the MNIST dataset, which is a derivative work from original NIST datasets, copyright by their authors and made available under the terms of the Creative Commons Attribution-Share Alike 3.0 license.