STARLING: Self-supervised Training of Text-based Reinforcement Learning Agent with Large Language Models.
Shreyas BasavatiaKeerthiram MurugesanShivam RatnakarPublished in: CoRR (2024)
Keyphrases
- language model
- learning agent
- reinforcement learning
- language modeling
- n gram
- probabilistic model
- information retrieval
- retrieval model
- query expansion
- language models for information retrieval
- state space
- test collection
- learning capabilities
- learning algorithm
- smoothing methods
- optimal policy
- expert systems
- relevance model
- learning process
- dynamic environments
- training set
- multimedia
- reinforcement learning algorithms
- neural network