Explainability-Performance Tradeoff

Definition

The observation that for deep neural networks there is an almost inverse relationship between performance and explainability: more layers and weights improve accuracy but eliminate high-level insight into why a decision was made.

Cast within

No topic-region cast yet — this appears once Explainability-Performance Tradeoff's compiled claims are aligned into a topic region's argument tree.

Claims mentioning Explainability-Performance Tradeoff (1)

Your notes

Loading notes...