Statements (33)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
gptkb:artificial_intelligence
|
| gptkbp:hasComponent |
hidden state
input layer output layer recurrent connection |
| gptkbp:hasProperty |
can process variable-length sequences
shares parameters across time steps |
| gptkbp:hasVariant |
gptkb:GRU
gptkb:LSTM |
| https://www.w3.org/2000/01/rdf-schema#label |
RNNs
|
| gptkbp:introduced |
gptkb:John_Hopfield
|
| gptkbp:introducedIn |
1980s
|
| gptkbp:limitation |
exploding gradient problem
vanishing gradient problem |
| gptkbp:openSource |
gptkb:TensorFlow
gptkb:Keras gptkb:PyTorch |
| gptkbp:relatedTo |
convolutional neural networks
feedforward neural networks |
| gptkbp:replacedBy |
gptkb:Transformers
|
| gptkbp:standsFor |
Recurrent Neural Networks
|
| gptkbp:trainer |
gptkb:backpropagation_through_time
|
| gptkbp:usedFor |
machine translation
speech recognition time series prediction text generation sequence modeling language modeling |
| gptkbp:usedIn |
natural language processing
music generation video analysis |
| gptkbp:bfsParent |
gptkb:Transformer_models
|
| gptkbp:bfsLayer |
7
|