adaumsilva/XAI-framework

Batch explanations: explain many rows at once and summarise them 30 days ago

enhancement good first issue

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Documentation site (mkdocs-material) with API reference about 1 month ago

documentation help wanted

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Integration tests for XGBoost, LightGBM and CatBoost about 1 month ago

help wanted good first issue testing

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Tutorial notebook: reading a consensus explanation about 1 month ago

documentation good first issue

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Add Explanation.to_markdown() for notebooks and reports about 1 month ago

enhancement good first issue

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Fast exact Shapley values for tree ensembles about 1 month ago

enhancement help wanted explainer performance

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Integrated Gradients for PyTorch / TensorFlow models about 1 month ago

enhancement help wanted deep-learning explainer

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Conformance test suite for third-party explainers about 1 month ago

help wanted testing

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

When methods disagree, say why: disagreement diagnostics about 1 month ago

enhancement help wanted research

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Release automation: publish to PyPI on tag via trusted publishing about 1 month ago

help wanted infrastructure

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Performance: explainer caching, parallel permutation importance, benchmarks about 1 month ago

help wanted performance

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Standalone HTML report with interactive chart about 1 month ago

enhancement help wanted

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Command-line interface: xai explain model.pkl data.csv --row 3 about 1 month ago

enhancement help wanted

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Image models: superpixel surrogate and coalition explanations about 1 month ago

enhancement help wanted explainer image

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Text models: surrogate and coalition explanations over tokens about 1 month ago

enhancement help wanted explainer text

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Surrogate: quantile discretisation, forward selection, and parity tests against reference implementations about 1 month ago

enhancement help wanted explainer

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Explain sklearn Pipelines in the original feature space about 1 month ago

enhancement help wanted

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Faithfulness & stability metrics, and metric-weighted consensus about 1 month ago

enhancement help wanted research

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Global explainers: Partial Dependence and Accumulated Local Effects about 1 month ago

enhancement help wanted explainer

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Add a counterfactual explainer ("what would need to change?") about 1 month ago

enhancement help wanted explainer

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Add an Anchors explainer (rule-based local explanations) about 1 month ago

enhancement help wanted explainer

adaumsilva/XAI-framework

1

XAI-Framework is a model-agnostic explainable-AI library with its own implementations of the ideas behind the most popular attribution methods

Python

Open Issues Need Help