An Agentic Approach to Generating XAI-Narratives

2026-03-20

AI Summary

This work addresses the gap in existing explainable AI (XAI) methods, which predominantly cater to experts and lack natural language explanations accessible to general users. We propose the first multi-agent collaborative framework for XAI narrative generation, where a Narrator produces an initial explanation and multiple Critic Agents iteratively refine it based on faithfulness and coherence criteria. The framework employs an iterative optimization process combined with a majority-vote ensemble strategy to enhance narrative quality. We instantiate five types of LLM-based agents—Basic, Critic, Critic-Rule, Coherent, and Coherent-Rule—and evaluate them across five tabular datasets and leading large language models. Experimental results demonstrate that Critic-based designs substantially improve narrative faithfulness; notably, under the Basic configuration, Claude-4.5-Sonnet reduces unfaithful content by 90% after three iterations, and the ensemble strategy further boosts performance across four models.

Technology Category

• Humans and AI: Explainable AI (XAI) for Human Understanding
• Multiagent Systems: Adversarial Agents
• Natural Language Processing: Generation

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Agentic search
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Explainable AI (XAI) research has experienced substantial growth in recent years. Existing XAI methods, however, have been criticized for being technical and expert-oriented, motivating the development of more interpretable and accessible explanations. In response, large language model (LLM)-generated XAI narratives have been proposed as a promising approach for translating post-hoc explanations into more accessible, natural-language explanations. In this work, we propose a multi-agent framework for XAI narrative generation and refinement. The framework comprises the Narrator, which generates and revises narratives based on feedback from multiple Critic Agents on faithfulness and coherence metrics, thereby enabling narrative improvement through iteration. We design five agentic systems (Basic Design, Critic Design, Critic-Rule Design, Coherent Design, and Coherent-Rule Design) and systematically evaluate their effectiveness across five LLMs on five tabular datasets. Results validate that the Basic Design, the Critic Design, and the Critic-Rule Design are effective in improving the faithfulness of narratives across all LLMs. Claude-4.5-Sonnet on Basic Design performs best, reducing the number of unfaithful narratives by 90% after three rounds of iteration. To address recurrent issues, we further introduce an ensemble strategy based on majority voting. This approach consistently enhances performance for four LLMs, except for DeepSeek-V3.2-Exp. These findings highlight the potential of agentic systems to produce faithful and coherent XAI narratives.

Problem

• Explainable AI
• XAI narratives
• faithfulness
• coherence
• natural-language explanations

Innovation

• multi-agent framework
• XAI narratives
• faithfulness
• iterative refinement
• large language models