Semi-Offline Reinforcement Learning for Optimized Text Generation.
Changyu ChenXiting WangYiqiao JinVictor Ye DongLi DongJie CaoYi LiuRui YanPublished in: ICML (2023)
Keyphrases
- text generation
- reinforcement learning
- natural language generation
- natural language
- function approximation
- reinforcement learning algorithms
- state space
- theorem prover
- real time
- robotic control
- learning process
- learning problems
- model free
- multi agent
- markov decision process
- learning algorithm
- policy search
- optimal policy
- action selection
- reinforcement learning methods
- multi agent reinforcement learning
- genetic algorithm