coq-neural-net-interp-computed.opam

opam-version: "2.0"
authors: [
  "Jason Gross <jgross@mit.edu>"
]
maintainer: "Jason Gross <jgross@mit.edu>"
homepage: "https://github.com/JasonGross/neural-net-coq-interp"
bug-reports: "https://github.com/JasonGross/neural-net-coq-interp/issues"
license: "MIT"
build: [make "-j%{jobs}%" "computed"]
install: []
depends: [
  "coq" {>= "8.17~" & ( < "8.18~" | > "8.18.0" )}
]
conflict-class: [
  "coq-neural-net-interp"
]
dev-repo: "git+https://github.com/JasonGross/neural-net-coq-interp.git"
synopsis: "Work in progress formalization of mechanistic interpretability arguments on tiny neural nets (only native-compiler tests)"
description: """
This opam package is mostly for testing the native compiler on Coq's
bench using a work-in-progress formalization of mechanistic
interpretability results.
"""
tags: ["logpath:NeuralNetInterp"]
url {
  src: "git+https://github.com/JasonGross/neural-net-coq-interp.git#main"
}