Saved in:
Bibliographic Details
Main Authors: Sibaee, Serry, Nacar, Omer, Ammar, Adel, Al-Habashi, Yasser, Al-Batati, Abdulrahman, Boulila, Wadii
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2506.01920
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912409686376448
author Sibaee, Serry
Nacar, Omer
Ammar, Adel
Al-Habashi, Yasser
Al-Batati, Abdulrahman
Boulila, Wadii
author_facet Sibaee, Serry
Nacar, Omer
Ammar, Adel
Al-Habashi, Yasser
Al-Batati, Abdulrahman
Boulila, Wadii
contents This paper addresses critical gaps in Arabic language model evaluation by establishing comprehensive theoretical guidelines and introducing a novel evaluation framework. We first analyze existing Arabic evaluation datasets, identifying significant issues in linguistic accuracy, cultural alignment, and methodological rigor. To address these limitations in LLMs, we present the Arabic Depth Mini Dataset (ADMD), a carefully curated collection of 490 challenging questions spanning ten major domains (42 sub-domains, see Figure 1. Using ADMD, we evaluate five leading language models: GPT-4, Claude 3.5 Sonnet, Gemini Flash 1.5, CommandR 100B, and Qwen-Max. Our results reveal significant variations in model performance across different domains, with particular challenges in areas requiring deep cultural understanding and specialized knowledge. Claude 3.5 Sonnet demonstrated the highest overall accuracy at 30\%, showing relative strength in mathematical theory in Arabic, Arabic language, and islamic domains. This work provides both theoretical foundations and practical insights for improving Arabic language model evaluation, emphasizing the importance of cultural competence alongside technical capabilities.
format Preprint
id arxiv_https___arxiv_org_abs_2506_01920
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle From Guidelines to Practice: A New Paradigm for Arabic Language Model Evaluation
Sibaee, Serry
Nacar, Omer
Ammar, Adel
Al-Habashi, Yasser
Al-Batati, Abdulrahman
Boulila, Wadii
Computation and Language
This paper addresses critical gaps in Arabic language model evaluation by establishing comprehensive theoretical guidelines and introducing a novel evaluation framework. We first analyze existing Arabic evaluation datasets, identifying significant issues in linguistic accuracy, cultural alignment, and methodological rigor. To address these limitations in LLMs, we present the Arabic Depth Mini Dataset (ADMD), a carefully curated collection of 490 challenging questions spanning ten major domains (42 sub-domains, see Figure 1. Using ADMD, we evaluate five leading language models: GPT-4, Claude 3.5 Sonnet, Gemini Flash 1.5, CommandR 100B, and Qwen-Max. Our results reveal significant variations in model performance across different domains, with particular challenges in areas requiring deep cultural understanding and specialized knowledge. Claude 3.5 Sonnet demonstrated the highest overall accuracy at 30\%, showing relative strength in mathematical theory in Arabic, Arabic language, and islamic domains. This work provides both theoretical foundations and practical insights for improving Arabic language model evaluation, emphasizing the importance of cultural competence alongside technical capabilities.
title From Guidelines to Practice: A New Paradigm for Arabic Language Model Evaluation
topic Computation and Language
url https://arxiv.org/abs/2506.01920