AI Glossary
The complete dictionary of Artificial Intelligence
Transparency Audit
Systematic and independent evaluation aimed at verifying the level of transparency of an AI model, including documentation of its data, algorithms, and decision-making processes.
Compliance Report
Formalized document resulting from an audit, attesting to an AI system's compliance with regulations and standards regarding transparency, fairness, and reliability.
Data Lineage
Complete tracking of the lifecycle of data used to train and evaluate a model, from their origin to their transformation, essential for transparency auditing and reproducibility.
Surrogate Model
Simple and interpretable model (such as a decision tree) trained to mimic the behavior of a complex black-box model, in order to provide a global explanation of its functioning.
XAI by Design
Methodological approach integrating interpretability and transparency constraints from the model design phase, rather than adding them a posteriori.
Explanation Confidence Score
Metric quantifying the degree of certainty that a given explanation faithfully represents the underlying decision logic of the model for a specific prediction.