/explore
|
You are here 
|
www.huffpost.com | ||
| | | | |
kffhealthnews.org
|
|
| | | | | Advocates of cheap and widely available vaccines thought the pandemic might change business as usual. They were wrong. | |
| | | | |
pluralistic.net
|
|
| | | | | ||
| | | | |
theintercept.com
|
|
| | | | | Urging Biden's reversal, industry lobbyists instruct Congress to talk up loss of jobs and fear of China. | |
| | | | |
www.paepper.com
|
|
| | | Today's paper: Rethinking 'Batch' in BatchNorm by Wu & Johnson BatchNorm is a critical building block in modern convolutional neural networks. Its unique property of operating on "batches" instead of individual samples introduces significantly different behaviors from most other operations in deep learning. As a result, it leads to many hidden caveats that can negatively impact model's performance in subtle ways. This is a citation from the paper's abstract and the emphasis is mine which caught my attention. Let's explore these subtle ways which can negatively impact your model's performance! The paper of Wu & Johnson can be found on arxiv. | ||
