Article, Early Access,
On the Opacity of Deep Neural Networks
Affiliations
- [1] Univ Copenhagen, Ctr Philosophy AI, Copenhagen, Denmark [NORA names: KU University of Copenhagen; University; Denmark; Europe, EU; Nordic; OECD]
Abstract
Abstract not displayed. As this article is not marked as Open Access, it is unclear if we are allowed to show the abstract. Please use the link in the sidebar to view the data provider version of the article including abstract.
Keywords
deep neural networks,
explainability,
mitigation,
model size,
opacity
