{"slug": "ai-elites-at-nadi-2026-shared-task-selecting-on-out-of-domain-accuracy-for-cross", "title": "AI Elites at NADI 2026 Shared Task: Selecting on Out-of-Domain Accuracy for Cross-Domain Spoken Arabic Dialect Identification", "summary": "The AI Elites team placed fourth in the NADI-2026 Subtask 2 country-level spoken Arabic dialect identification shared task, reaching 91.54% in-domain, 73.10% on the held-out MADIS-5 corpus (scored five-way), and 52.51% on the blind test set (C_avg 0.11), according to the team's paper in the Proceedings of the Fourth Arabic Natural Language Processing Conference. The team selected on out-of-domain accuracy at every decision point, training the chosen Arabic ASR encoder on ADI20-micro mixed with Casablanca; that criterion picked an encoder in-domain accuracy would have rejected by 32 points on a held-out corpus, a gap more than twenty times the measured seed variation. Of the 39.03-point in-domain-to-blind gap, at least 18.44 points are attributable to leaving the training domain, and the authors trace part of the remainder to a corpus that entered the training mixture while still being used for checkpoint selection.", "body_md": "##### Abstract\n\nWe describe our submission to NADI-2026 Subtask 2, country-level spoken Arabic dialect identification on a blind test set from undisclosed sources. Because the target domain is unknown at development time, we select on out-of-domain accuracy rather than in-domain validation accuracy at every decision point, comparing three backbones under a common pooling-plus-linear head and training the selected Arabic ASR encoder on ADI20-micro mixed with Casablanca. The system reaches 91.54% in-domain, 73.10% on the held-out MADIS-5 corpus (scored five-way), and 52.51% on the blind test set (C\n<sub>avg</sub>0.11), placing fourth. The criterion helped, choosing an encoder that in-domain accuracy would have rejected by 32 points on a held-out corpus, a gap more than twenty times our measured seed variation, but it was not sufficient: of the 39.03-point in-domain to blind gap, at least 18.44 points are attributable to leaving the training domain. We trace part of the remainder to a corpus that entered the training mixture while still being used for checkpoint selection.\n\n- Anthology ID:\n- 2026.arabicnlp-sharedtasks.10\n- Volume:\n- [Proceedings of The Fourth Arabic Natural Language Processing Conference: Shared Tasks](https://aclanthology.org/volumes/2026.arabicnlp-sharedtasks/)\n- Month:\n- October\n- Year:\n- 2026\n- Address:\n- Budapest, Hungary\n- Editors:\n- [Muhammad Abdul-Mageed](https://aclanthology.org/people/muhammad-abdul-mageed/) ,[Tamer Elsayed](https://aclanthology.org/people/tamer-elsayed/) ,[Houda Bouamor](https://aclanthology.org/people/houda-bouamor/unverified/) ,[Samhaa El-Beltagy](https://aclanthology.org/people/samhaa-r-el-beltagy/) ,[Hend Al-Khalifa](https://aclanthology.org/people/hend-al-khalifa/) ,[Ahmed Abdelali](https://aclanthology.org/people/ahmed-abdelali/unverified/) ,[AbdelRahim Elmadany](https://aclanthology.org/people/abdelrahim-elmadany/unverified/) ,[Abdellah El Mekki](https://aclanthology.org/people/abdellah-el-mekki/) ,[Kareem Darwish](https://aclanthology.org/people/kareem-darwish/) ,[Sakhar Alkhereyf](https://aclanthology.org/people/sakhar-alkhereyf/) ,[Hamzah Luqman](https://aclanthology.org/people/hamzah-luqman/) ,[Salam Khalifa](https://aclanthology.org/people/salam-khalifa/) ,[Samar Magdy](https://aclanthology.org/people/samar-magdy/unverified/) ,[Zaid Alyafeai](https://aclanthology.org/people/zaid-alyafeai/) ,[Injy Hamed](https://aclanthology.org/people/injy-hamed/unverified/) ,[Wajdi Zaghouani](https://aclanthology.org/people/wajdi-zaghouani/) ,[Imed Zitouni](https://aclanthology.org/people/imed-zitouni/unverified/)\n- Venue:\n- [ArabicNLP](https://aclanthology.org/venues/arabicnlp/)\n- SIG:\n- [SIGARAB](https://aclanthology.org/sigs/sigarab/)\n- Publisher:\n- Association for Computational Linguistics\n- Note:\n- Pages:\n- 88–94\n- Language:\n- URL:\n- [https://aclanthology.org/2026.arabicnlp-sharedtasks.10/](https://aclanthology.org/2026.arabicnlp-sharedtasks.10/)\n- DOI:\n- Cite (ACL):\n- Musab Iskandar, Mohammed Al-Haig, Adel Sameer Alharbi, Khalid Alharbi, and Youssef Mohamed. 2026. [AI Elites at NADI 2026 Shared Task: Selecting on Out-of-Domain Accuracy for Cross-Domain Spoken Arabic Dialect Identification](https://aclanthology.org/2026.arabicnlp-sharedtasks.10/) . In*Proceedings of The Fourth Arabic Natural Language Processing Conference: Shared Tasks* , pages 88–94, Budapest, Hungary. Association for Computational Linguistics.\n- Cite (Informal):\n- [AI Elites at NADI 2026 Shared Task: Selecting on Out-of-Domain Accuracy for Cross-Domain Spoken Arabic Dialect Identification](https://aclanthology.org/2026.arabicnlp-sharedtasks.10/) (Iskandar et al., ArabicNLP 2026)\n- PDF:\n- [https://aclanthology.org/2026.arabicnlp-sharedtasks.10.pdf](https://aclanthology.org/2026.arabicnlp-sharedtasks.10.pdf)", "url": "https://wpnews.pro/news/ai-elites-at-nadi-2026-shared-task-selecting-on-out-of-domain-accuracy-for-cross", "canonical_source": "https://aclanthology.org/2026.arabicnlp-sharedtasks.10/", "published_at": "2026-10-23 00:00:00+00:00", "updated_at": "2026-10-24 09:43:37.636340+00:00", "lang": "en", "topics": ["natural-language-processing", "machine-learning", "ai-research"], "entities": ["AI Elites", "NADI 2026", "Musab Iskandar", "Mohammed Al-Haig", "Adel Sameer Alharbi", "Khalid Alharbi", "Youssef Mohamed", "Association for Computational Linguistics"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/ai-elites-at-nadi-2026-shared-task-selecting-on-out-of-domain-accuracy-for-cross", "markdown": "https://wpnews.pro/news/ai-elites-at-nadi-2026-shared-task-selecting-on-out-of-domain-accuracy-for-cross.md", "text": "https://wpnews.pro/news/ai-elites-at-nadi-2026-shared-task-selecting-on-out-of-domain-accuracy-for-cross.txt", "jsonld": "https://wpnews.pro/news/ai-elites-at-nadi-2026-shared-task-selecting-on-out-of-domain-accuracy-for-cross.jsonld"}}