Conversational XAI
Conversational XAI aims to improve user understanding of AI decision-making processes through interactive dialogue, moving beyond static explanations. Current research focuses on developing frameworks for managing user requests, adapting explanations based on user feedback and understanding, and employing various explanation methods (e.g., narrative-based, logic-based) within conversational agents, often leveraging large language models. This approach holds significant promise for enhancing trust and transparency in AI systems across diverse applications, particularly by enabling more effective communication of complex information to non-expert users.
Papers
August 13, 2024
June 12, 2024
May 16, 2024
March 21, 2024
September 29, 2023
May 16, 2023