chore: start fork by removing Python code and experiments

.github/workflows/CI-Julia.yml → .github/workflows/CI.yml

@@ -1,4 +1,4 @@
-name: CI-Julia
+name: CI
 on:
   push:
     branches:
@@ -7,43 +7,41 @@ on:
   pull_request:
   workflow_dispatch:
 concurrency:
+  # Skip intermediate builds: always.
+  # Cancel intermediate builds: only if it is a pull request build.
   group: ${{ github.workflow }}-${{ github.ref }}
   cancel-in-progress: ${{ startsWith(github.ref, 'refs/pull/') }}
 jobs:
   test:
-    name: Julia ${{ matrix.version }} - ${{ matrix.os }} - ${{ matrix.arch }}
-    runs-on: ${{ matrix.os }}
+    name: Julia ${{ matrix.version }} (${{ matrix.group }})
+    runs-on: ubuntu-latest
     timeout-minutes: 60
-    permissions:
+    permissions: # needed to allow julia-actions/cache to proactively delete old caches that it has created
       actions: write
       contents: read
     strategy:
       fail-fast: false
       matrix:
         version:
           - 'lts'
-        os:
-          - ubuntu-latest
-        arch:
-          - x64
+          - '1'
+
     steps:
-      - uses: actions/checkout@v5
-      - uses: julia-actions/setup-julia@v2
+      - uses: actions/checkout@v7
+      - uses: julia-actions/setup-julia@v3
         with:
           version: ${{ matrix.version }}
-          arch: ${{ matrix.arch }}
-      - uses: julia-actions/cache@v2
-      - name: Install dependencies & run tests
-        run: julia --project=./julia --color=yes -e '
-          using Pkg;
-          Pkg.test("SmoothedDifferentiation"; coverage=true);'
+          arch: x64
+      - uses: julia-actions/cache@v3
+      - uses: julia-actions/julia-buildpkg@v1
+      - uses: julia-actions/julia-runtest@v1
       - uses: julia-actions/julia-processcoverage@v1
         with:
-          directories: ./julia/src
-      - uses: codecov/codecov-action@v5
+          directories: src,ext
+      - uses: codecov/codecov-action@v7
         with:
-          files: lcov.info
           token: ${{ secrets.CODECOV_TOKEN }}
+          files: lcov.info
           fail_ci_if_error: false
   runic:
     name: Runic formatting

.gitignore

@@ -2,179 +2,5 @@
 *.jl.*.cov
 *.jl.cov
 *.jl.mem
-/julia/Manifest.toml
-/julia/test/Manifest.toml
+*/Manifest.toml
 settings.json
-
-# Byte-compiled / optimized / DLL files
-__pycache__/
-*.py[cod]
-*$py.class
-
-# C extensions
-*.so
-
-# Distribution / packaging
-.Python
-build/
-develop-eggs/
-dist/
-downloads/
-eggs/
-.eggs/
-lib/
-lib64/
-parts/
-sdist/
-var/
-wheels/
-share/python-wheels/
-*.egg-info/
-.installed.cfg
-*.egg
-MANIFEST
-
-# PyInstaller
-#  Usually these files are written by a python script from a template
-#  before PyInstaller builds the exe, so as to inject date/other infos into it.
-*.manifest
-*.spec
-
-# Installer logs
-pip-log.txt
-pip-delete-this-directory.txt
-
-# Unit test / coverage reports
-htmlcov/
-.tox/
-.nox/
-.coverage
-.coverage.*
-.cache
-nosetests.xml
-coverage.xml
-*.cover
-*.py,cover
-.hypothesis/
-.pytest_cache/
-cover/
-*.lock
-
-# Translations
-*.mo
-*.pot
-
-# Django stuff:
-*.log
-local_settings.py
-db.sqlite3
-db.sqlite3-journal
-
-# Flask stuff:
-instance/
-.webassets-cache
-
-# Scrapy stuff:
-.scrapy
-
-# Sphinx documentation
-docs/_build/
-
-# PyBuilder
-.pybuilder/
-target/
-
-# Jupyter Notebook
-.ipynb_checkpoints
-
-# IPython
-profile_default/
-ipython_config.py
-
-# pyenv
-#   For a library or package, you might want to ignore these files since the code is
-#   intended to run in multiple environments; otherwise, check them in:
-# .python-version
-
-# pipenv
-#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
-#   However, in case of collaboration, if having platform-specific dependencies or dependencies
-#   having no cross-platform support, pipenv may install dependencies that don't work, or not
-#   install all needed dependencies.
-#Pipfile.lock
-
-# UV
-#   Similar to Pipfile.lock, it is generally recommended to include uv.lock in version control.
-#   This is especially recommended for binary packages to ensure reproducibility, and is more
-#   commonly ignored for libraries.
-#uv.lock
-
-# poetry
-#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
-#   This is especially recommended for binary packages to ensure reproducibility, and is more
-#   commonly ignored for libraries.
-#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
-#poetry.lock
-
-# pdm
-#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
-#pdm.lock
-#   pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
-#   in version control.
-#   https://pdm.fming.dev/latest/usage/project/#working-with-version-control
-.pdm.toml
-.pdm-python
-.pdm-build/
-
-# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
-__pypackages__/
-
-# Celery stuff
-celerybeat-schedule
-celerybeat.pid
-
-# SageMath parsed files
-*.sage.py
-
-# Environments
-.env
-.venv
-env/
-venv/
-ENV/
-env.bak/
-venv.bak/
-
-# Spyder project settings
-.spyderproject
-.spyproject
-
-# Rope project settings
-.ropeproject
-
-# mkdocs documentation
-/site
-
-# mypy
-.mypy_cache/
-.dmypy.json
-dmypy.json
-
-# Pyre type checker
-.pyre/
-
-# pytype static type analyzer
-.pytype/
-
-# Cython debug symbols
-cython_debug/
-
-# PyCharm
-#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
-#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
-#  and can be added to the global gitignore or merged into this file.  For a more nuclear
-#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
-#.idea/
-
-# PyPI configuration file
-.pypirc

README.md

@@ -1,56 +1,5 @@
-# SmoothDiff
+# SmoothDiff.jl
 
-Code for the NeurIPS 2025 paper *"Smoothed Differentiation Efficiently Mitigates Shattered Gradients in Explanations"*.
+Julia reference implementation of SmoothDiff for the NeurIPS 2025 paper *"Smoothed Differentiation Efficiently Mitigates Shattered Gradients in Explanations"*.
 
-## Code
-
-* The Julia reference implementation of SmoothDiff can be found in the [`/julia`](/julia) folder, which contains a Julia package called `SmoothedDifferentiation.jl`.
-* The PyTorch reference implementation of SmoothDiff can be found in [`/python`](/python) folder.
-
-## Installation
-
-1. [Install Julia](https://julialang.org/downloads/) `v1.11`. On Unix systems, this requires running
-    ```bash
-    curl -fsSL https://install.julialang.org | sh
-    ```
-2. Start a [Julia REPL](https://docs.julialang.org/en/v1/stdlib/REPL/) session by typing `julia` in your terminal
-3. Install [DrWatson.jl](https://github.com/JuliaDynamics/DrWatson.jl) by typing the following in your Julia REPL:
-    ```julia-repl
-    ]add DrWatson
-    ```
-4. Run the experiments and plotting scripts listed below by typing `include("path/to/file.jl")` in your REPL, replacing the string with the correct path.
-
-## Experiments
-
-> [!NOTE]
-> The experiment code refers to the state of the repository at the time of publication.
-> See the [`neurips25`](https://github.com/adrhill/smoothdiff-experiments/tree/neurips25) tag for the exact code used in the paper.
-
-We provide all code and virtual environments required to reproduce our experiments and figures.
-
-### Running experiments
-
-The following steps need to be run in sequence:
-
-1. [Download the ImageNet dataset](https://image-net.org/download.php) by agreeing to its terms of access and run the script [`/experiments/heatmaps/save_input.jl`](/experiments/heatmaps/save_input.jl) to save preprocessed input tensors.
-2. Compute explanations by running [`/experiments/cluster/run_analyzers/run.jl`](/experiments/cluster/run_analyzers/run.jl) as well as the $n=10^6$ sample SmoothGrad run in [`/experiments/cluster/run_analyzers/run_100k.jl`](/experiments/cluster/run_analyzers/run_100k.jl).
-3. Compute benchmarks by running [`/experiments/cluster/run_analyzers/run.jl`](/experiments/cluster/run_analyzers/run_benchmarks.jl).
-4. After having computed explanations in step 3, evaluate them using pixel-flipping by running [`/experiments/cluster/pixelflipping/run.jl`](/experiments/cluster/pixelflipping/run.jl).
-5. The raw quantitative results from Appendix F can be found in [`/experiments/cluster/quantus-eval/results`](/experiments/cluster/quantus-eval/results) and are computed by running [`/experiments/cluster/quantus-eval/run.sh`](/experiments/cluster/quantus-eval/run.sh).
-
-### Reproducing figures
-
-After having computed explanations and pixel-flipping results, the following plots and tables can be reproduced by running the respective file:
-
-- **Figure 1 (a)**: create data by running [`/experiments/shattered_gradient_1d/run_fig_1a.jl`](/experiments/shattered_gradient_1d/run_fig_1a.jl), then plot using [`/experiments/shattered_gradient_1d/plot_fig_1a.jl`](/experiments/shattered_gradient_1d/plot_fig_1a.jl) 
-- **Figure 1 (b)**: run [`/experiments/heatmaps/grid_fig_1b.jl`](/experiments/heatmaps/grid_fig_1b.jl)
-- **Figure 2**: run [`/experiments/ssim_convergence/convergence_ssim_combined.jl`](/experiments/ssim_convergence/convergence_ssim_combined.jl)
-- **Figure 3** and **Figure 15**: run [`/experiments/heatmaps/grid_convergence.jl`](/experiments/heatmaps/grid_convergence.jl)
-- **Figure 4** and **Table 1**: run [`/experiments/pixelflipping_plots/srg_plot.jl`](/experiments/pixelflipping_plots/srg_plot.jl)
-- **Figure 5**: run [`/experiments/colormaps/jetvsbatlow.jl`](/experiments/colormaps/jetvsbatlow.jl)
-- **Figure 6**: run [`/experiments/heatmaps/grid_pipeline_comparison.jl`](/experiments/heatmaps/grid_pipeline_comparison.jl)
-- **Figure 7**: run [`/experiments/ssim_convergence/convergence_cosine_sim.jl`](/experiments/ssim_convergence/convergence_cosine_sim.jl)
-- **Figures 8 to 13**: run [`/experiments/cluster/quantus-eval/plot.jl`](/experiments/cluster/quantus-eval/plot.jl)
-- **Figure 14**: run [`/experiments/beta-smoothing/plot_beta_smoothing.jl`](/experiments/beta-smoothing/plot_beta_smoothing.jl)
-- **Figure 16**: run [`/experiments/heatmaps/grid_multiclass.jl`](/experiments/heatmaps/grid_multiclass.jl)
-- **Figure 17**: run [`/experiments/heatmaps/grid_appendix.jl`](/experiments/heatmaps/grid_appendix.jl)
+The full experiments can be found here: https://github.com/adrhill/smoothdiff-experiments/