Explainable AI

Methods and tools that make AI model decisions transparent, interpretable and verifiable to support trust, compliance and debugging.

Explainable AI (XAI) comprises techniques for representing and assessing the decision basis of machine learning models.

Maturity

Emerging

Cognitive loadHigh

360° overview

Six perspectives place the building block in context. The numbers show where each perspective continues in the reading path.

The building block at a glance

Explainable AI

360°

2

2

Integrations

TensorFlow and PyTorch models

+2

1

Transparency: Explanations should be traceable and reproducible.

Value stream stage: Build

4

3

4