AI RESEARCH
RPS: Information Elicitation with Reinforcement Prompt Selection
arXiv CS.LG
•
ArXi:2604.13817v1 Announce Type: new Large language models (LLMs) have shown remarkable capabilities in dialogue generation and reasoning, yet their effectiveness in eliciting user-known but concealed information in open-ended conversations remains limited. In many interactive AI applications, such as personal assistants, tutoring systems, and legal or clinical, users often withhold sensitive or uncertain information due to privacy concerns, ambiguity, or social hesitation. This makes it challenging for LLMs to gather complete and contextually relevant inputs.