.png){kind=icon}
Return
Deconstructing the generalization gap
.png){kind=icon}
PRE
AI
DOI：10.1038/s42256-023-00766-7.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
New research reveals a duality between neural network weights and neuron activities that enables a geometric decomposition of the generalization gap. The framework provides a way to interpret the effects of regularization schemes such as stochastic gradient descent and dropout on generalization - and to improve upon these methods.
Journal
IF:
23.9
Papers:
1.3K
Citations:
1.5W
.png){kind=icon}
