|
gptkbp:instanceOf
|
gptkb:artificial_intelligence
|
|
gptkbp:addressingMechanism
|
differentiable addressing
|
|
gptkbp:application
|
algorithmic tasks
sequence learning
copy task
sorting task
|
|
gptkbp:citation
|
high
|
|
gptkbp:component
|
controller
read head
write head
memory matrix
|
|
gptkbp:controls
|
gptkb:feedforward_neural_network
recurrent neural network
|
|
gptkbp:developedBy
|
gptkb:DeepMind
|
|
gptkbp:field
|
gptkb:artificial_intelligence
gptkb:machine_learning
neural computation
|
|
gptkbp:goal
|
learn algorithms from data
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
Neural Turing Machine
|
|
gptkbp:influenced
|
differentiable neural computer
memory-augmented neural networks
|
|
gptkbp:inspiredBy
|
Technical Machine
|
|
gptkbp:introducedIn
|
2014
|
|
gptkbp:inventedBy
|
gptkb:Alex_Graves
gptkb:Greg_Wayne
gptkb:Ivo_Danihelka
|
|
gptkbp:memoryAccess
|
content-based addressing
location-based addressing
|
|
gptkbp:memoryType
|
external memory
|
|
gptkbp:notableFeature
|
differentiable memory access
combines neural networks with memory
|
|
gptkbp:notablePublication
|
Neural Turing Machines, Graves et al., 2014
|
|
gptkbp:openSource
|
gptkb:TensorFlow
gptkb:PyTorch
|
|
gptkbp:publishedIn
|
arXiv:1410.5401
|
|
gptkbp:relatedTo
|
differentiable computer
differentiable neural computer
|
|
gptkbp:trainer
|
gradient descent
|
|
gptkbp:type
|
end-to-end differentiable
|
|
gptkbp:bfsParent
|
gptkb:convolutional_neural_network
|
|
gptkbp:bfsLayer
|
5
|