GaussianDropout Layer
Layer AIO:GaussianDropoutLayer NOT_REVIEWED
A regularization layer that applies multiplicative 1-centered Gaussian noise.
Ontology IRI:
https://w3id.org/aio/GaussianDropoutLayer
Ontology Structure & Relationships
Parent Classes
Branches
Annotations & Sources
Definition Sources:
Subsets:
https://w3id.org/aio/LayerSubset
Curated AIMech Concepts Grounded to this Class
1 Concept| Curated Concept | Kind | Mapping Relation | Rationale |
|---|---|---|---|
Fast Dropout Training
AIMech:FastDropoutTraining |
Algorithms | related_match | Both involve Gaussian approximations to multiplicative dropout effects, but fast dropout also includes moment integration and variance propagation; it is not identical to the pinned training-only layer that injects one-centered Gaussian noise. |
Import Provenance & Upstream Synchronization
Source Repo: AIO GitHub
Release Version:
aio.jsonCommit:
397851c547…License: CC-BY-4.0
Imported At: 2026-10-10T07:34:10Z
SHA256:
8f741d71408c…
Source File:
data/aio/GaussianDropoutLayer.yaml