Tag
This paper introduces a tri-agent framework for evaluating and aligning the question clarification capabilities of large language models, using three LLM-based agents to simulate and assess clarification dialogues.