CARLA: A Python Library to Benchmark Algorithmic Recourse and Counterfactual Explanation Algorithms

A collection of algorithms of counterfactual explanations.

This repository contains the code and scripts to reproduce the results from our paper "HARE: Human-in-the-Loop Algorithmic Recourse", accepted at TMLR 2025.

GlassAlpha is an open-source toolkit for deterministic, regulator-ready ML audit reports. One command generates comprehensive, reproducible audits with TreeSHAP explainability, fairness metrics, calibration analysis, and robustness testing for XGBoost, LightGBM, and logistic regression models (binary classification, tabular data only).

An efficient + provably robust implementation of algorithmic recourse via Rashomon sets in PyTorch & Scikit-Learn (NeurIPS '25 Spotlight)

Replay-validated, path-based counterfactual explanations for LightGBM binary classifiers.