RLPrompt: Optimizing Discrete Text Prompts with Reinforcement Learning

31206

Knowledge Triples

Source Entity	Relationship	Target Entity
Few Shot Classification Task	isA	Classification Task Task
RLPrompt Method	appliesTo	BERT Method
RLPrompt Method	appliesTo	Classification Task Task
RLPrompt Method	appliesTo	Generation Task Task
RLPrompt Method	appliesTo	GPTs Material
RLPrompt Method	improves	Soft Prompts Method
RLPrompt Method	incorporates	Reward Stabilization Method
RLPrompt Method	optimizes	Discrete Prompts OtherEntity
Unsupervised Text Style Transfer Task	isA	Generation Task Task