bonXAI 🪴🌱🌿🍃

This repository contains the bonXAI demo library for the CHI 2024 workshop paper Explainability in JupyterLab and Beyond: Interactive XAI Systems for Integrated and Collaborative Workflows

Getting Started with bonXAI

The quickest way to ensure that bonXAI is installed correctly is to start with a clean conda environment with the exact versions of the following packages:

conda create -n bonxai python=3.8 jupyterlab=3.4 ipywidgets=7.7.5 ipykernel=5.3 pandas umap-learn pytorch torchvision

conda activate bonxai

Install pip for this new environment:

conda install pip

Install other required packages. Note that we do this separately as more recent versions of the datasets package has a breaking change in fsspec.

pip install datasets==2.16

Then install bonXAI:

pip install bonxai

The package should show up when you run:

jupyter labextension list

Finally, run the notebooks/sequence_classification.ipynb file to see how the bonXAI visualization modules can be used in a text classification workflow.

Troubleshooting

If you encounter errors when importing bonXAI in JupyterLab, first ensure that the package is successfully installed and appears in the Jupyter labextension list.

jupyter labextension list

If this has been verified, check that the python version used by JupyterLab is identical to the version in which bonXAI is installed. In cases where there are multiple virtual environments in the same machine, the bonXAI package may be installed in a different location.

Citation

If this work has been useful to you, please consider citing us as:

@article{guo2024explainability,
  title={Explainability in JupyterLab and Beyond: Interactive XAI Systems for Integrated and Collaborative Workflows},
  author={Guo, Grace and Arendt, Dustin and Endert, Alex},
  journal={https://arxiv.org/abs/2404.02081},
  year={2024}
}