QIAS 2025 Shared Task

 

 

 

We are pleased to announce the launch of the QIAS 2025 Shared Task, Question and Answer in Islamic STUDIES Assessment Shared Task, a research challenge dedicated to evaluating Large Language Models (LLMs) in the domain of Islamic knowledge.

Subtasks

Participants are invited to engage with two focused evaluation subtasks:

  • Islamic Inheritance Reasoning (ʿIlm al-Mawārīth)
  • Islamic Knowledge Assessment (covering topics such as Fiqh, Ḥadīth, Tafsīr, Uṣūl al-Fiqh, and more)

Each task features expertly validated multiple-choice questions (MCQs) categorized into three difficulty levels:

  • Beginner
  • Intermediate
  • Advanced

Resources Provided

Participants will have access to:

  • Baseline models
  • Training and evaluation datasets
  • Evaluation code and submission guidelines

Key Dates

  • Training Data Release: June 1, 2025
  • Test Set Release: July 25, 2025
  • Results Announcement: July 30, 2025
  • Workshop Presentation: ArabicNLP @ EMNLP 2025 (November 5–9, Suzhou, China)

Learn More

Visit the official website for detailed guidelines and participation instructions

 https://sites.google.com/view/qias2025 

We invite the research community to join us in advancing the frontiers of Arabic and Islamic AI through QIAS 2025.