Incognition logo
Home
Learn
Practice
Leaderboard
0
VisualizationsMachine Learning in Code

Model Interpretability

Click the dots to flip through four reasons a model's decisions need explaining: someone is owed an answer, catching mistakes, trust, and the law.

Someone is owed an answer
A person refused a loan can reasonably ask why, and "the model said so" is not a reason.

A model with millions of weights has no readable rule inside it. That is fine for a photo filter and not fine for a mortgage decision.

Customize