Technology term
Out-of-Distribution Generalization (OOD generalization)
A model's ability to perform reliably on data that differs meaningfully from the distribution represented in its training set. It tests whether the model learned transferable patterns rather than relying only on correlations specific to familiar conditions.
English
Out-of-Distribution Generalization
Arabic
التعميم خارج نطاق بيانات التدريب
First appeared in
This Robot Folded 785 Loads of Laundry. The Real Breakthrough Was Doing It in Unseen Homes
