Graphical Introduction to eXplainable AI Techniques for Space Applications

This document is a graphical introduction and summary of the Catalogue of Explainability Techniques for AI Applications in Space (XAI Catalogue) created as a part of the "PINEBERRY: Explainable, Robust and Secure AI for Demystifying Space Operations" activity funded by the European Space Agency. The full XAI catalogue is accessible for registered users from ESA Member States on the Space CODEV platform. The graphical introduction is a user-friendly educational material for anyone struggling with low trustworthiness of AI applications and a lack of transparency in their decisions. It offers an overview of all types of methods (i.e., attributions, sensitivity analysis, counterfactuals, and interpretable models) for different data modalities in the space domain (i.e., tabular, time series, natural language, and images). More information about the project can be found at its website: https://assurance-ai.space-codev.org/