Skip to content

Latest commit

 

History

History
27 lines (22 loc) · 1.01 KB

Nomenclature.md

File metadata and controls

27 lines (22 loc) · 1.01 KB

Nomenclature

Since recently, started approximately following the nomenclature used by Torch neural network package, since the names appear to me to be:

  • concise
  • simple
  • easy to understand

Concretely, the names for some things, nouns are:

  • input: the input to the layer, coming from previous layer
  • output: the output from the layer, going to the next layer
  • gradInput: the partial gradient of the loss with respect to the input to the layer
  • gradOutput: the partial gradient of the loss with respect to the output of the layer
  • gradWeights: partial gradient of the loss with respect to the weights (renaming not done yet)

Some verbs are:

  • forward: calculate the output, given the input
  • backward: calculate gradInput, given gradOutput

(Note: these names are a work in progress, there are still plenty of old names about, like:

  • errorsForUpstream => gradInput
  • outputFromUpstream => input
  • errors => gradOutput
  • propagate => forward
  • backprop => backward
  • etc ... )