Language-Conditioned Reinforcement Learning to Solve Misunderstandings with Action Corrections.
Frank RöderManfred EppePublished in: CoRR (2022)
Keyphrases
- reinforcement learning
- natural language
- action selection
- action space
- learning algorithm
- reward shaping
- temporal difference
- programming language
- learning problems
- function approximation
- language processing
- state space
- multi agent
- markov decision processes
- language learning
- optimal control
- model free
- reinforcement learning algorithms
- representation language
- transition model
- genetic algorithm