Studying the Inductive Biases of RNNs with Synthetic Variations of Natural Languages.
Shauli RavfogelYoav GoldbergTal LinzenPublished in: NAACL-HLT (1) (2019)
Keyphrases
- recurrent neural networks
- natural language
- context sensitive
- machine learning
- inductive learning
- natural language understanding
- real images are presented
- echo state networks
- semantic interpretation
- feedforward neural networks
- knowledge representation
- real time
- learning algorithm
- real world
- inductive bias
- face images
- evolutionary algorithm
- hidden layer
- fuzzy neural network
- knowledge base
- view angle
- neural network
- inductive reasoning