| dbo:thumbnail
| |
| dbp:caption
|
- Halo: each iteration on dimension partially shares input tensor values with the previous iteration. This occurs because iterating also indexes through a sum with an index on . (en)
- Stationarity: iterating on dimension continues to reuse the same input tensor values. (en)
|
| dbp:class
| |
| dbp:direction
| |
| dbp:header
|
- Examples of Data Reuse in Convolutions (en)
|
| dbp:image
|
- Halo Reuse in a Convolution.png (en)
- Stationarity Reuse in Convolution.png (en)
|
| dbp:width
| |
| dbp:wikiPageUsesTemplate
| |
| dct:subject
| |
| rdf:type
| |
| rdfs:label
|
- Spatial architecture (en)
|
| prov:wasDerivedFrom
| |
| foaf:depiction
| |
| foaf:isPrimaryTopicOf
| |
| is foaf:primaryTopic
of | |
