Transformer Feed-Forward Layers Are Key-Value Memories.
Mor GevaRoei SchusterJonathan BerantOmer LevyPublished in: EMNLP (1) (2021)
Keyphrases
- feed forward
- back propagation
- artificial neural networks
- neural network
- neural nets
- recurrent neural networks
- visual cortex
- error back propagation
- feed forward neural networks
- activation function
- fuzzy logic
- associative memory
- multi layer
- hidden layer
- biologically plausible
- neural architecture
- recurrent networks
- artificial intelligence
- single layer
- bp neural network
- fault diagnosis
- spiking neurons
- high voltage
- real time