Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

XAI for All: Can Large Language Models Simplify Explainable AI?

Philip Mavrepis, Georgios Makridis, Georgios Fatouros, Vasileios Koukos, Maria Margarita Separdani, Dimosthenis Kyriazis

TL;DR

This work introduces x-[plAIn], a GPT-based, audience-adaptive XAI explainer built with GPT-Builder to translate complex XAI outputs into concise, user-specific narratives. By integrating multiple XAI methods (e.g., LIME, SHAP, Grad-CAM) and employing prompting strategies with Chain-of-Thought reasoning, the system targets both non-expert end users and AI developers, aiming to accelerate understanding and decision-making. The authors validate the approach through five cross-domain use cases and a user survey, demonstrating favorable reception and the potential to bridge the gap between advanced AI explanations and practical applications. The study highlights a practical path toward inclusive, human-centered XAI and outlines future enhancements around customization granularity, visualization-textual integration, and bias monitoring.

Abstract

The field of Explainable Artificial Intelligence (XAI) often focuses on users with a strong technical background, making it challenging for non-experts to understand XAI methods. This paper presents "x-[plAIn]", a new approach to make XAI more accessible to a wider audience through a custom Large Language Model (LLM), developed using ChatGPT Builder. Our goal was to design a model that can generate clear, concise summaries of various XAI methods, tailored for different audiences, including business professionals and academics. The key feature of our model is its ability to adapt explanations to match each audience group's knowledge level and interests. Our approach still offers timely insights, facilitating the decision-making process by the end users. Results from our use-case studies show that our model is effective in providing easy-to-understand, audience-specific explanations, regardless of the XAI method used. This adaptability improves the accessibility of XAI, bridging the gap between complex AI technologies and their practical applications. Our findings indicate a promising direction for LLMs in making advanced AI concepts more accessible to a diverse range of users.

XAI for All: Can Large Language Models Simplify Explainable AI?

TL;DR

This work introduces x-[plAIn], a GPT-based, audience-adaptive XAI explainer built with GPT-Builder to translate complex XAI outputs into concise, user-specific narratives. By integrating multiple XAI methods (e.g., LIME, SHAP, Grad-CAM) and employing prompting strategies with Chain-of-Thought reasoning, the system targets both non-expert end users and AI developers, aiming to accelerate understanding and decision-making. The authors validate the approach through five cross-domain use cases and a user survey, demonstrating favorable reception and the potential to bridge the gap between advanced AI explanations and practical applications. The study highlights a practical path toward inclusive, human-centered XAI and outlines future enhancements around customization granularity, visualization-textual integration, and bias monitoring.

Abstract

The field of Explainable Artificial Intelligence (XAI) often focuses on users with a strong technical background, making it challenging for non-experts to understand XAI methods. This paper presents "x-[plAIn]", a new approach to make XAI more accessible to a wider audience through a custom Large Language Model (LLM), developed using ChatGPT Builder. Our goal was to design a model that can generate clear, concise summaries of various XAI methods, tailored for different audiences, including business professionals and academics. The key feature of our model is its ability to adapt explanations to match each audience group's knowledge level and interests. Our approach still offers timely insights, facilitating the decision-making process by the end users. Results from our use-case studies show that our model is effective in providing easy-to-understand, audience-specific explanations, regardless of the XAI method used. This adaptability improves the accessibility of XAI, bridging the gap between complex AI technologies and their practical applications. Our findings indicate a promising direction for LLMs in making advanced AI concepts more accessible to a diverse range of users.
Paper Structure (25 sections, 8 figures, 1 table)

This paper contains 25 sections, 8 figures, 1 table.

Table of Contents

  1. Introduction
  2. Background
  3. eXplainable AI (XAI)
  4. From explainability to interpretability
  5. Challenges in Communicating AI Concepts
  6. Objectives
  7. Literature Review
  8. Explainable AI: A Technical Overview
  9. Language Models in AI
  10. Large Language Models in XAI
  11. Methodology
  12. Role of GPT-Builder in LLM Development
  13. Use cases
  14. Use Case 1
  15. Use Case 2
  16. ...and 10 more sections

Figures (8)

  • Figure 1: Taxonomy of XAI Methods
  • Figure 2: Plot for SHAPley values evaluation in makridis2022evaluating.
  • Figure 3: Plot for LIME evaluation of fake news from szczepanski2021new.
  • Figure 4: Plot for Saliency Map Verbalization (SMV) from feldhus2023saliency.
  • Figure 5: Plot for Grad-CAM explanations from moujahid2022combining.
  • ...and 3 more figures