Author

W. Zaghouani

h-index 29 View on Semantic Scholar

We have 3 of 162 papers

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Preprint Aug 2026

When Do VLMs Help Arabic Manuscript OCR? A Cross-Dataset Study

Vision-language models (VLMs) are increasingly being used for document understanding, yet their role in Arabic and Islamic manuscript recognition remains underexplored. To address such a gap in this paper, we evaluate traditional OCR, general-purpose VLMs, Arabic-specialized VLMs, and OCR-conditioned VLM correction acr...

M. Farazi, Firoj Alam, A. Maaradji et al. · 0 citations

#natural language process... Preprint Aug 2026

ArabicDialectSafety: A Dialect-Aware Benchmark for Arabic Content Safety Classification

This paper introduces a dual-task evaluation framework for binary safe/unsafe detection and granular harm classification across dialects, and evaluates seven frontier LLMs as response generators on harmful dialectal Arabic prompts and observes unsafe generation rates below 5 percent across models.

Wajdi Zaghouani, Mukut Biswas, K. Aldous et al. · 0 citations

Open access 2026

Lost in Dialect: The Annotation Gap in Multilingual LLM Safety

The MSAG is formalized as a framework for analyzing systematic weaknesses in multilingual safety annotation pipelines, identifying four sources of bias: language coverage gaps, dialect representation gaps, cultural semantic gaps, and annotation guideline gaps.

W. Zaghouani · Proceedings of the 1st Works... · 0 citations