Logical Activation Functions: Logit-space equivalents of Probabilistic Boolean Operators.
Scott C. LoweRobert EarleJason d'EonThomas TrappenbergSageev OorePublished in: NeurIPS (2022)
Keyphrases
- activation function
- boolean operators
- neural network
- neural architecture
- extended boolean
- feed forward
- artificial neural networks
- probabilistic model
- bayesian networks
- input space
- basis functions
- back propagation
- radial basis function
- low dimensional
- principal component analysis
- web pages
- decision making
- machine learning