Modular Representation Underlies Systematic Generalization in Neural Natural Language Inference Models.
Atticus GeigerKyle RichardsonChristopher PottsPublished in: CoRR (2020)
Keyphrases
- natural language
- distributed representations
- neural network model
- random fields
- network architecture
- probabilistic model
- statistical dependencies
- representational power
- complex systems
- neural network
- dynamic bayesian networks
- probabilistic modeling
- inference process
- semantic analysis
- semantic representation
- qualitative models
- model construction
- parameter estimation
- information extraction
- machine learning