The Ambiguity Mitigator develops knowledge-guided health AI agents that recognize when a question is incomplete, underspecified, or open to multiple interpretations. By asking targeted clarifying questions and grounding responses in structured medical knowledge, the project aims to make conversational health systems safer, more accurate, and more useful.

Selected Publications

• A Knowledge-Guided Agentic Framework for Mitigating Patient-Context Ambiguity in Health Queries
• HealthQ: Unveiling Questioning Capabilities of LLM Chains in Healthcare Conversations
• MedCoT-RAG: Causal Chain-of-Thought RAG for Medical Question Answering
• CDF-RAG: Causal Dynamic Feedback for Adaptive Retrieval-Augmented Generation