Explainable AI (XAI)

Explainable AI develops ways to make model behavior understandable. Interpretability describes how comprehensible a model is, while feature importance estimates which inputs influence its predictions.

Japanese terms

1. Explainable AI — 説明可能せつめいかのうなAI: Methods and practices intended to make AI behavior understandable to people.
2. Interpretability — 解釈可能性かいしゃくかのうせい: The degree to which a person can understand how a model produces an output.
3. Feature importance — 特徴量重要度とくちょうりょうじゅうようど: A measure of how strongly an input feature influences a model’s predictions.

Related topics

AI safety and evaluation uses explanations as one source of evidence, and AI ethics considers whether explanations support meaningful accountability.