gmanch94

explainability

gmanch94/claude-code-aidlc-template/.claude/skills/explainability/SKILL.md

ML model explainability — SHAP global/local explanations, LIME, partial dependence plots, counterfactuals, method selection by model type and audience. Use when asked about "model explainability", "why did the model predict X?", "feature importance", "SHAP", "LIME", "black box", "model interpretability", regulatory explanation requirements, or debugging model behavior.