Learning Interpretability Tool (LIT)
Explainable AI (XAI); Machine Learning
noun phrase
Definition: An interactive visual tool for understanding, debugging, and analysing machine-learning models across modalities such as text, images, and tabular data. In current usage, LIT is an extensible platform for model inspection, counterfactual testing, salience-based analysis, and dataset slicing [Drechsel, Herbold 2026].
Example in context: “We used the Learning Interpretability Tool (LIT) [30] to run and visualise the token-level importance …” [Sabuncuoglu et al. 2025]
Synonym: LIT
Related terms: interpretability tool; model debugging; XAI; counterfactual inspection; dataset slicing