Enhancing Reinforcement Learning with discrete interfaces to learn the Dyck Language.
Florian DietzDietrich KlakowPublished in: CoRR (2021)
Keyphrases
- reinforcement learning
- function approximators
- natural language
- learning agent
- function approximation
- language processing
- language learning
- state space
- programming language
- user interface
- continuous state spaces
- learning process
- action space
- discrete geometry
- discrete version
- object oriented
- database systems
- multi agent
- discrete space
- sensory inputs
- imitation learning
- reinforcement learning agents
- continuous state and action spaces