What AI says about Xplique

Xplique is a Python toolkit for explainable AI that aggregates state-of-the-art methods to help users understand complex neural networks. It offers attribution methods (e.g., Saliency, Grad-CAM, Integrated Gradients), feature visualization, concept-based explanations, and metrics, originally built for TensorFlow and with partial PyTorch support. The library also includes an Example-based module that retrieves relevant dataset examples (similar, counterfactuals, prototypes) and provides tutorials, a quick-start guide, and pip installation.