The 2nd Human-Centric eXplainable AI in Education (HEXED) Workshop

As machine learning models and their larger language model counterparts become more prevalent in education, the challenge of interpretability has grown alongside their adoption. While researchers have explored various approaches, often drawing from the broader eXplainable AI (XAI) community, current methods remain limited. We highlight the need to rethink explainability in educational settings. The 2nd iteration of the Human-Centric eXplainable AI in Education (HEXED) workshop brings together researchers dedicated to advancing interpretability in AI-driven education. Our goals are to (1) establish a shared vision and common vocabulary for XAI in education, (2) facilitate the exchange of recent research and best practices, (3) brainstorm practical methods to enhance model transparency that are specific to the education domain, and (4) define evaluation metrics for assessing explanations and interpretability with teachers, students, and parents. We also aim to discuss what the rise of LLMs means for the field of eXplainable AI and seek synergies to enhance LLMs with methods in XAI towards increasing student, parent, and teacher trust. Through research presentations and structured discussions, we aim to address key challenges and shape the future of explainable AI in education.