shap/shap

Health v2 · maintenance only

Adoption not part of the score

25783 stars · 3754 forks observed · 2026-09-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-10-12, confidence 0.98

SHAP (SHapley Additive exPlanations) is a Python library that provides game‑theoretic explanations for the predictions of any machine‑learning model. It implements fast exact and approximate algorithms for tree ensembles, linear models, deep networks, and model‑agnostic cases.

Use cases

• Explain predictions of a trained XGBoost or LightGBM model
• Generate feature importance plots for a scikit‑learn classifier
• Interpret individual predictions of a neural network for debugging
• Assess fairness and bias by visualizing contribution of sensitive features
• Create interactive explanations for a deployed model in a web app

When to choose

• You need mathematically sound, additive feature attributions based on Shapley values
• Your workflow already uses Python and common ML libraries like scikit‑learn, XGBoost, or PyTorch
• You require fast tree‑based explanations with optional GPU acceleration

When to avoid

• You need real‑time explanations on low‑power edge devices without Python support
• Your model is not supported by SHAP’s optimized algorithms and you cannot afford the computational cost of model‑agnostic sampling
• You prefer a purely visual, no‑code tool rather than a code library

Facets

library · maturity active

machine-learning data-visualization artificial-intelligence machine-learning deep-learning python windows shap model-interpretation game-theory feature-attribution explainability linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="shap/shap")