SIGN-XAI2 Documentation

SIGN (Sign-based Improvement of Gradient-based explaNations) is a novel XAI method intended to reduce bias in explanations that are intrinsically induced by several state-of-the-art XAI methods. The SIGN-XAI2 package enables simple application of this method in your projects and is based on Zennit and LRP for Transformers. If your are using TensorFlow instead of PyTorch, have a look at our TF-version of SIGN-XAI.

SIGN-based explanations are particularly well suited for generating bias-reduced heatmaps for both image and time series data, enhancing interpretability by more reliably uncovering relevant features.

Install

$ pip install signxai2

Contents

• Getting started
  • Install
  • Basic Usage
  • Example Scripts
• How-to
• Tutorials
• Rules
• Composites
• DFT-LRP

Citing

If you use this package or parts of it in your own work, please consider citing our paper:

@article{Gumpfer2023SIGN,
    title = {SIGNed explanations: Unveiling relevant features by reducing bias},
    author = {Nils Gumpfer and Joshua Prim and Till Keller and Bernhard Seeger and Michael Guckert and Jennifer Hannig},
    journal = {Information Fusion},
    pages = {101883},
    year = {2023},
    issn = {1566-2535},
    doi = {https://doi.org/10.1016/j.inffus.2023.101883}
}