Repository for replicating the ANCHOR evaluation results of the paper "Benchmarking and Survey of Explanation Methods for Black Box Models".
git clone https://github.com/DLemming/hsfd-data-mining
cd hsfd-data-miningWe use uv for dependency management (faster than pip). To install all dependencies and activate venv, run:
uv sync
source .venv/bin/activateUnlike the original paper, this will automatically install the correct versions of python and all dependencies.
Make sure you're in the project's root directory. The jupyter notebook repro_german_anchor.ipynb contains all the code that was used in the attempt to reproduce the original paper's results. Inside the notebook, active the .venv initialized by uv, and run experiments as you like!