This tutorial aims to explore the intersection be tween XAI and the field of Natural Language Processing (NLP) by highlighting its significance for various tasks. By combining both theoretical and practical aspects of XAI, we provide a comprehensive overview of its definitions and review the state-of-the-art methodologies employed to explain NLP models. While previous tutorials have focused on establishing the theoretical foundations of the XAI field and providing essential concepts to the audience, this tutorial seeks to bridge the gap between theory and practice. We will discuss and demonstrate the most commonly used XAI techniques, the prominent libraries applicable to different models, and the limitations of each method. Furthermore, we will highlight the methods that can be utilized to visualize these models while performing specific NLP tasks.
• Gain an understanding of the fundamental concepts of XAI and its importance: highlighting the need for XAI techniques from both theoretical and practical standpoints. It will also provide an introduction to Pre-trained Language Models (PLMs), summarize the prerequisites, and outline different angles covered in the subsequent sections of the tutorial.
• Explore the latest methodologies and techniques for explaining LLMs: where the participant will delve into the state-of-the-art explanation methods and visualization techniques that can enhance the interpretability of LLM.
• Apply these techniques to real-world data in various NLP tasks: through hands-on projects, participants will practice applying XAI techniques to real-world data, thereby gaining the practical skills of experimenting.
• Discuss the challenges, future directions, and gaps of LLM explainability: fostering a deeper understanding of the current limitations and potential advancements in the field.
The tutorial is designed to be accessible to a wide range of audience, providing valuable insights without requiring prior knowledge of the XAI field.
This tutorial targets:
• Scholars and researchers who focus on computational aspects of human languages and are interested in exploring the existing language models.
• PhD candidates and graduate students who want to incorporate XAI techniques with their projects to gain a deeper understanding of LLM.
• Linguists with a strong background in linguistics who are curious about the advancement of the models and the impact of such advancement on the language itself.
The outcome of this tutorial can be summarized as the following:
• Understanding XAI main concepts and categories.
• Apply many XAI techniques to explain the behavior of LLMs in NLP.
• Identify the gaps and potential advancement in XAI and current LLM.
This virtual tutorial will last three hours and cover the following:
• Introduction to Pre-trained Language Models (PLMs) (10 minutes)
• Overview and Brief History of Explainability in AI (10 minutes)
• Categorization of Explainability Techniques (20 minutes)
• In-depth Practical Sessions:
• Conclusion and Future Challenges in Explainable NLP (10 minutes)
• Q&A Session (10 minutes)
• Explainable Natural Language Modeling (Søgaard, 2021).
• Explainability for large language models: A survey (Zhao et al., 2024).
• Towards Faithful Model Explanation in NLP: A Survey (Lyu et al., 2024).
• Usable XAI: 10 Strategies Towards Exploiting Explainability in the LLM Era (Wu et al., 2024).
• A Primer on the Inner Workings of Transformer-based Language Models (Ferrando et al., 2024).
• Large Language Models Cannot Explain Themselves (Sarkar, 2024).
• The Explainability of Transformers: Current Status and Directions (Fantozzi and Naldi, 2024).
@article{Alrajhi2025Coling,
author = {Wafa Alrajhi, Nourah Alangari Sharefah Al-Ghamdi and Hend Al-Khalifa},
title = {From Theory to Practice: A Hands-on Tutorial in Explainable NLP},
journal = {COLING},
year = {2025},
}