---
title: "Grounding Arabic LLMs in the Doha Historical Dictionary — Eltanbouly 2026"
description: "Large language models, including Arabic-native ones, often misread Classical Arabic. 85% accuracy. 38 references."
authors: ["Somaya Eltanbouly", "Samer Rashwani"]
year: 2026
venue: "arXiv:2603.23972"
source: "https://arxiv.org/abs/2603.23972"
doi: "10.48550/arxiv.2603.23972"
arxiv: "2603.23972"
licence: "arXiv — non-exclusive license to distribute"
licence_url: "http://arxiv.org/licenses/nonexclusive-distrib/1.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/eltanbouly-2026-grounding-arabic-llms-doha-historical/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/eltanbouly-2026-grounding-arabic-llms-doha-historical/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/7/) / Grounding Arabic LLMs in the Doha Historical Dictionary: Retrieval-Augmented Understanding of Quran and Hadith

# Grounding Arabic LLMs in the Doha Historical Dictionary: Retrieval-Augmented Understanding of Quran and Hadith

**Authors:** Somaya Eltanbouly and Samer Rashwani · **Year:** 2026 · **Venue:** arXiv:2603.23972 · **Licence:** [arXiv — non-exclusive license to distribute](http://arxiv.org/licenses/nonexclusive-distrib/1.0/)

**Links:** [Publisher's page](https://arxiv.org/abs/2603.23972) · [PDF](https://arxiv.org/pdf/2603.23972) · [DOI 10.48550/arxiv.2603.23972](https://doi.org/10.48550/arxiv.2603.23972) · [arXiv 2603.23972](https://arxiv.org/abs/2603.23972) · [OpenAlex](https://openalex.org/W7140858648)

**Citations:** 0 citations (OpenAlex, 2026-10-14)

## Summary
Large language models, including Arabic-native ones, often misread Classical Arabic. The Qur'an and Hadith use archaic vocabulary, senses that drifted over centuries, and older grammar. Models trained mostly on modern web text therefore produce anachronistic or wrong interpretations. Earlier dictionary-grounded RAG work for Arabic relied only on a modern lexicon.

The authors instead ground generation in the Doha Historical Dictionary of Arabic (DHDA), which traces how word meanings developed and backs each stage with dated citations. Their pipeline works in four steps:
- It cleans each query and weights its key terms.
- It predicts what kind of information the user wants.
- It retrieves dictionary entries with BM25 plus dense embeddings and reorders them with a fine-tuned cross-encoder.
- It sends the evidence to the LLM through a prompt template picked by the predicted intent.

With this setup, Fanar and ALLaM exceed 85% accuracy and come close to Gemini. Gemini also acts as the automatic judge, and its verdicts were checked against human annotation. The study shows that a diachronic lexicon can serve as a working knowledge base for historical and religious text understanding.

## Key points
- The knowledge source is the Doha Historical Dictionary of Arabic (DHDA). Code and resources are released at github.com/somayaeltanbouly/Doha-Dictionary-RAG.
- Intent detection uses TF-IDF features fed to a Random Forest. The categories it predicts cover a word's etymology, its first usage, its basic and contextual meanings, the author, the date, the inscription, a list of derivations, morphology, the source, terminological usage, and a special "Quranic first usage" class. Predictions below a confidence threshold fall into "other".
- Retrieval is hybrid. BM25 handles exact matches on headwords, roots and citation phrases. Four multilingual embedding models, chosen from the Arabic RAG leaderboard, were compared for the dense side. A fine-tuned cross-encoder re-ranks the results.
- Fanar and ALLaM pass 85% accuracy with RAG. Gemini serves as LLM-as-a-judge, and its judgments agreed with human evaluation at κ = 0.87.
- The error analysis points to three main problem areas:
  - Linguistic: diacritics, compound expressions, and ambiguous dictionary entries.
  - Retrieval: misses of relevant documents.
  - Instruction following: Fanar sometimes drifted into Chinese text, and the smaller models less often admitted that the retrieved documents did not answer the question.

## Interesting ideas
- The system classifies the question's intent first and then uses a prompt template specific to that intent, with a low-confidence fallback to a generic path; this keeps each answer type handled by a focused procedure.
- Exact lexical matching is paired with similarity-based retrieval because dictionary queries hinge on precise headwords and roots that fuzzy matching may lose.
- An LLM judge is trusted only after its agreement with human raters is measured (κ = 0.87), rather than being assumed reliable.
- A dictionary that records senses with dates and first attestations lets the system interpret old texts by period-appropriate meanings instead of modern ones.
- Models are explicitly told to say when the evidence does not contain the answer, and the authors check whether each model obeys.

## For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

- **I1** (answer): School-test and fable QA readers could detect the question type (who, why, what happened first, what a word means) and dispatch to a dedicated reader module, falling back to a general reader when the classifier is unsure.
- **I2** (world): Lookups into WordNet/Wikidata/the knowledge graph could try exact lemma and sense-key matches first and use similarity only as a logged secondary candidate channel, so every retrieved fact remains traceable.
- **I3** (learn): Before an LLM teacher tier grades silver data or answers at scale, compute its agreement with a small human-checked sample and gate its use on that score.
- **I4** (world): Attaching period or register tags to word senses in the compiled ontology would help when reading older fables and tales, where archaic senses would otherwise be resolved to their modern meaning.
- **I5** (answer): Readers should abstain when the text's world lacks supporting facts, and abstention behaviour can be tested with negative-control questions whose answers are deliberately absent.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

Fields: [Linguistics & languages](https://md.research.elamur.ai/free/en/shelves/linguistics/)

Topics: [🧠 Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · [🐪 Semitic languages](https://md.research.elamur.ai/free/en/topics/semitic-languages/) · [🔍 Information retrieval & search](https://md.research.elamur.ai/free/en/topics/information-retrieval/)

## Cite

Cite as: Somaya Eltanbouly and Samer Rashwani. 2026. Grounding Arabic LLMs in the Doha Historical Dictionary: Retrieval-Augmented Understanding of Quran and Hadith. arXiv:2603.23972.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@misc{eltanbouly2026grounding,
  title = {{Grounding Arabic LLMs in the Doha Historical Dictionary: Retrieval-Augmented Understanding of Quran and Hadith}},
  author = {Somaya Eltanbouly and Samer Rashwani},
  year = {2026},
  doi = {10.48550/arxiv.2603.23972},
  eprint = {2603.23972},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2603.23972}
}
```

- [← previous GreekCommonGen: A Benchmark for Evaluating Generative Commonsense Reasoning in Greek](https://md.research.elamur.ai/free/en/papers/stamopoulos-2026-greekcommongen-benchmark-evaluating-generative-commonsense/)
- [next → HalleluBERT: Let Every Token That Has Meaning Bear Its Weight](https://md.research.elamur.ai/free/en/papers/schmitt-2026-hallelubert-let-every-token-that/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/7/)

## References

38 references, 11 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Mashaabi, M., Al-Khalifa, S., Al-Khalifa, H.: A Survey of Large Language Models for Arabic Language and its Dialects (2025). https://arxiv.org/abs/2410.20238 [↗](https://arxiv.org/abs/2410.20238) [our card](https://md.research.elamur.ai/free/en/papers/zhao-2023-survey-large-language-models/)
2. Al-Khalifa, S., Durrani, N., Al-Khalifa, H., Alam, F.: The landscape of arabic large language models. Commun. ACM 68 (10), 54–61 (2025) https://doi.org/10.1145/3737453 [↗](https://doi.org/10.1145/3737453)
3. Nacar, O., Sibaee, S.T., Ahmed, S., Ben Atitallah, S., Ammar, A., Alhabashi, Y., Al-Batati, A.S., Alsehibani, A., Qandos, N., Elshehy, O., Abdelkader, M., Koubaa, A.: Towards inclusive Arabic LLMs: A culturally aligned benchmark in Arabic large language model evaluation. In: Hettiarachchi, H., Ranasinghe, T., Rayson, P., Mitkov, R., Gaber, M., Premasiri, D., Tan, F.A., Uyangodage, L. (eds.) Proceedings of the First Workshop on Language Models for Low-Resource Languages, pp. 387–401. Association for Computational Linguistics, Abu Dhabi, United Arab Emirates (2025). https://aclanthology.org/2025.loreslm-1.29/ [↗](https://aclanthology.org/2025.loreslm-1.29/)
4. Inoue, G., Alhafni, B., Baimukan, N., Bouamor, H., Habash, N.: The interplay of variant, size, and task type in Arabic pre-trained language models. In: Proceedings of the Sixth Arabic Natural Language Processing Workshop. Association for Computational Linguistics, Kyiv, Ukraine (Online) (2021) [our card](https://md.research.elamur.ai/free/en/papers/inoue-2021-interplay-variant-size-task-type/)
5. Malhas, R., Elsayed, T.: Arabic machine reading comprehension on the holy qur’an using cl-arabert. Information Processing & Management 59 (6), 103068 (2022) https://doi.org/10.1016/j.ipm.2022.103068 [↗](https://doi.org/10.1016/j.ipm.2022.103068)
6. Sengupta, N., Sahu, S.K., Jia, B., Katipomu, S., Li, H., Koto, F., Marshall, W., Gosal, G., Liu, C., Chen, Z., Afzal, O.M., Kamboj, S., Pandit, O., Pal, R., Pradhan, L., Mujahid, Z.M., Baali, M., Han, X., Bsharat, S.M., Aji, A.F., Shen, Z., Liu, Z., Vassilieva, N., Hestness, J., Hock, A., Feldman, A., Lee, J., Jackson, A., Ren, H.X., Nakov, P., Baldwin, T., Xing, E.: Jais and Jais-chat: Arabic-Centric Foundation and Instruction-Tuned Open Generative Large Language Models (2023). https://arxiv.org/abs/2308.16149 [↗](https://arxiv.org/abs/2308.16149) [our card](https://md.research.elamur.ai/free/en/papers/sengupta-2023-jais-jais-chat-arabic-centric/)
7. Almazrouei, E., Alobeidli, H., Alshamsi, A., Cappelli, A., Cojocaru, R., Debbah, M., Goffinet, Hesslow, D., Launay, J., Malartic, Q., Mazzotta, D., Noune, B., Pannier, B., Penedo, G.: The Falcon Series of Open Language Models (2023). https://arxiv.org/abs/2311.16867 [↗](https://arxiv.org/abs/2311.16867) [our card](https://md.research.elamur.ai/free/en/papers/almazrouei-2023-falcon-series-open-language-models/)
8. silma-ai: SILMA 9B Instruct SILMA-9B-Instruct-v1.0 (2024) [find](https://scholar.google.com/scholar?q=silma%2Dai%3A+SILMA+9B+Instruct+SILMA%2D9B%2DInstruct%2Dv1%2E0+%282024%29)
9. Bari, M.S., Alnumay, Y., Alzahrani, N.A., Alotaibi, N.M., Alyahya, H.A., AlRashed, S., Mirza, F.A., Alsubaie, S.Z., Alahmed, H.A., Alabduljabbar, G., Alkhathran, R., Almushayqih, Y., Alnajim, R., Alsubaihi, S., Mansour, M.A., Hassan, S.A., Alrubaian, D.M., Alammari, A., Alawami, Z., Al-Thubaity, v1.0. [https://huggingface.co/silma-ai/A](https://huggingface.co/silma-ai/A)., Abdelali, A., Kuriakose, J., Abujabal, A., Al-Twairesh, N., Alowisheq, A., Khan, H.: ALLam: Large language models for arabic and english. In: The Thirteenth International Conference on Learning Representations (2025). https://openreview.net/forum?id=MscdsFVZrN [↗](https://huggingface.co/silma-ai/A) [our card](https://md.research.elamur.ai/free/en/papers/bari-2024-allam-large-language-models-arabic/)
10. Team, F., Abbas, U., Ahmad, M.S., Alam, F., Altinisik, E., Asgari, E., Boshmaf, Y., Boughorbel, S., Chawla, S., Chowdhury, S., Dalvi, F., Darwish, K., Durrani, N., Elfeky, M., Elmagarmid, A., Eltabakh, M., Fatehkia, M., Fragkopoulos, A., Hasanain, M., Hawasly, M., Husaini, M., Jung, S.-G., Lucas, J.K., Magdy, W., Messaoud, S., Mohamed, A., Mohiuddin, T., Mousi, B., Mubarak, H., Musleh, A., Naeem, Z., Ouzzani, M., Popovic, D., Sadeghi, A., Sencar, H.T., Shinoy, M., Sinan, O., Zhang, Y., Ali, A., Kheir, Y.E., Ma, X., Ruan, C.: Fanar: An Arabic-Centric Multimodal Generative AI Platform (2025). https://arxiv.org/abs/2501.13944 [↗](https://arxiv.org/abs/2501.13944) [our card](https://md.research.elamur.ai/free/en/papers/abbas-2025-fanar-arabic-centric-multimodal-generative/)
11. Anwar, M., Freihat, A., Ibrahim, G., Awad, M., Sadallah, A.A.M.A., Gosal, G., Ramakrishnan, G., Chandran, S., Mishra, B., Joshi, R., Frikha, A., Goffinet, E., Maiti, A., El Filali, A., Al Barri, S., Ghosh, S., Pal, R., Mullah, P., Shukla, A., Siddiki, S., Kamboj, S., Pandit, O., Sahu, S., El Badawy, A., Mohamed, A., Chamma, A., Dufraisse, E., Bounhar, A., Bouch, D., Abdine, H., Shang, G., Koto, F., Wang, Y., Xie, Z., Mekky, A., Elbadry, R.H.E., Ahmad, S., Ahsan, M., El-Herraoui, O.E.M., Orel, D., Iqbal, H., Elzeky, K.M.N.A.F., Abassy, M., Ali, K., Eletter, S., Atif, F., Mukhituly, N., Li, H., Han, X., Singh, A., Quraishi, Z., Sengupta, N., Murray, L., Sheinin, A., Hestness, J., Vassilieva, N., Ren, H., Liu, Z., Vazirgiannis, M., Nakov, P.: Jais 2: A family of Arabic-centric open large language models. Technical report, IFM (December 2025) [our card](https://md.research.elamur.ai/free/en/papers/anwar-2026-jais-2-family-arabic-centric/)
12. Karnak: A Depth-Extended Arabic-English LLM. Applied Innovation Center (2026) [find](https://scholar.google.com/scholar?q=Karnak%3A+A+Depth%2DExtended+Arabic%2DEnglish+LLM%2E+Applied+Innovation+Center+%282026%29)
13. Alghallabi, W., Thawkar, R., Ghaboura, S., More, K., Thawakar, O., Cholakkal, H., Khan, S., Anwer, R.M.: Fann or Flop: A Multigenre, Multiera Benchmark for Arabic Poetry Understanding in LLMs (2025). https://arxiv.org/abs/2505.18152 [↗](https://arxiv.org/abs/2505.18152)
14. Al-Rasheed, R., Al Muaddi, A., Aljasim, H., Al-Matham, R., Alhoshan, M., Al Wazrah, A., AlOsaimy, A.: Evaluating RAG pipelines for Arabic lexical information retrieval: A comparative study of embedding and generation models. In: El-Haj, M. (ed.) Proceedings of the 1st Workshop on NLP for Languages Using Arabic Script, pp. 155–164. Association for Computational Linguistics, Abu Dhabi, UAE (2025). https://aclanthology.org/2025.abjadnlp-1.16/ [↗](https://aclanthology.org/2025.abjadnlp-1.16/)
15. Kutuzov, A., Øvrelid, L., Szymanski, T., Velldal, E.: Diachronic word embeddings and semantic shifts: a survey. In: Bender, E.M., Derczynski, L., Isabelle, P. (eds.) Proceedings of the 27th International Conference on Computational Linguistics, pp. 1384–1397. Association for Computational Linguistics, Santa Fe, New Mexico, USA (2018). https://aclanthology.org/C18-1117/ [↗](https://aclanthology.org/C18-1117/) [our card](https://md.research.elamur.ai/free/en/papers/kutuzov-2018-diachronic-word-embeddings-semantic-shifts/)
16. Degaetano-Ortlieb, S., Fonteyn, L., Krielke, P., Teich, E.: Workshop: Computational models of diachronic language change organizers. Technical report (2023) [find](https://scholar.google.com/scholar?q=Degaetano%2DOrtlieb%2C+S%2E%2C+Fonteyn%2C+L%2E%2C+Krielke%2C+P%2E%2C+Teich%2C+E%2E%3A+Workshop%3A+Computational+models+of+diachronic+language+change+organizers%2E+Technical+report+%282023%29)
17. Theron, R., Fontanillo, L.: Diachronic-information visualization in historical dictionaries. Information Visualization 14 , 111–136 (2015) https://doi.org/10.1177/1473871613495844 [↗](https://doi.org/10.1177/1473871613495844)
18. Schlechtweg, D., Zamora-Reina, F.D., Bravo-Marquez, F., Arefyev, N.: Sense through time: diachronic word sense annotations for word sense induction and lexical semantic change detection. Language Resources and Evaluation 59 (2), 1431–1465 (2025) https://doi.org/10.1007/s10579-024-09771-7 [↗](https://doi.org/10.1007/s10579-024-09771-7)
19. Hammo, B., Yagi, S., Ismail, O., AbuShariah, M.: Exploring and exploiting a historical corpus for arabic. Language Resources and Evaluation 50 (4), 839–861 (2016) https://doi.org/10.1007/s10579-015-9304-9 [↗](https://doi.org/10.1007/s10579-015-9304-9)
20. Boudelaa, S.: Aralex: a lexical database for modern standard arabic. Behavior research methods 42 , 481–7 (2010) https://doi.org/10.3758/BRM.42.2.481 [↗](https://doi.org/10.3758/brm.42.2.481)
21. Attia, M., Pecina, P., Toral, A., Tounsi, L., Genabith, J.: An open-source finite state morphological transducer for Modern Standard Arabic. In: Maletti, A., Constant, M. (eds.) Proceedings of the 9th International Workshop on Finite State Methods and Natural Language Processing, pp. 125–133. Association for Computational Linguistics, Blois, France (2011). https://aclanthology.org/W11-4417/ [↗](https://aclanthology.org/W11-4417/) [our card](https://md.research.elamur.ai/free/en/papers/attia-2011-open-source-finite-state-morphological/)
22. Miyagawa, S.: RAG-enhanced neural machine translation of Ancient Egyptian text: A case study of THOTH AI. In: Hämäläinen, M., Öhman, E., Bizzoni, Y., Miyagawa, S., Alnajjar, K. (eds.) Proceedings of the 5th International Conference on Natural Language Processing for Digital Humanities, pp. 33–40. Association for Computational Linguistics, Albuquerque, USA (2025). https://doi.org/10.18653/v1/2025.nlp4dh-1.4 . https://aclanthology.org/2025.nlp4dh-1.4/ [↗](https://doi.org/10.18653/v1/2025.nlp4dh-1.4)
23. Li, J., Wei, T., Qu, W., Li, B., Feng, M., Wang, D.: Combining lexicon definitions and the retrieval-augmented generation of a large language model for the automatic annotation of ancient chinese poetry. Mathematics 13 (2025) https://doi.org/10.3390/math13122023 [↗](https://doi.org/10.3390/math13122023)
24. Tran, T.T., González-Gallardo, C.E., Doucet, A.: Retrieval augmented generation for historical newspapers. In: Proceedings of the ACM/IEEE Joint Conference on Digital Libraries. Institute of Electrical and Electronics Engineers Inc., ??? (2025). https://doi.org/10.1145/3677389.3702542 [↗](https://doi.org/10.1145/3677389.3702542)
25. Kosem, I., Jakubicek, M., Kallas, J., Krek, S.: Electronic Lexicography in the 21st Century Linking Lexical Data in the Digital Age: Proceedings of eLex 2015 Conference, 11-13 August 2015, Herstmonceux Castle, United Kingdom. Trojina, Ljubljana (2015). OCLC: 938563540 [find](https://scholar.google.com/scholar?q=Kosem%2C+I%2E%2C+Jakubicek%2C+M%2E%2C+Kallas%2C+J%2E%2C+Krek%2C+S%2E%3A+Electronic+Lexicography+in+the+21st+Century+Linking+Lexical+Data+in+the+Digital+Age%3A+Proceedings+of+eLex+2015+Conference%2C+11%2D13+August+2015%2C+Herstmonceu)
26. Schryver, G.M.: Generative ai and lexicography: The current state of the art using chatgpt. International Journal of Lexicography 36 , 355–387 (2023) https://doi.org/10.1093/ijl/ecad021 [↗](https://doi.org/10.1093/ijl/ecad021)
27. Robertson, S.E., Walker, S.: Some simple effective approximations to the 2-poisson model for probabilistic weighted retrieval. In: Proceedings of the 17th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval. SIGIR ’94, pp. 232–241. Springer, Berlin, Heidelberg (1994) [find](https://scholar.google.com/scholar?q=Robertson%2C+S%2EE%2E%2C+Walker%2C+S%2E%3A+Some+simple+effective+approximations+to+the+2%2Dpoisson+model+for+probabilistic+weighted+retrieval%2E+In%3A+Proceedings+of+the+17th+Annual+International+ACM+SIGIR+Conference+on+)
28. Sturua, S., Mohr, I., Akram, M.K., Günther, M., Wang, B., Krimmel, M., Wang, F., Mastrapas, G., Koukounas, A., Koukounas, A., Wang, N., Xiao, H.: jina-embeddings-v3: Multilingual Embeddings With Task LoRA (2024). https://arxiv.org/abs/2409.10173 [↗](https://arxiv.org/abs/2409.10173)
29. Chen, J., Xiao, S., Zhang, P., Luo, K., Lian, D., Liu, Z.: BGE M3-Embedding: Multi-Lingual, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillation (2024) [find](https://scholar.google.com/scholar?q=Chen%2C+J%2E%2C+Xiao%2C+S%2E%2C+Zhang%2C+P%2E%2C+Luo%2C+K%2E%2C+Lian%2C+D%2E%2C+Liu%2C+Z%2E%3A+BGE+M3%2DEmbedding%3A+Multi%2DLingual%2C+Multi%2DFunctionality%2C+Multi%2DGranularity+Text+Embeddings+Through+Self%2DKnowledge+Distillation+%282024%29)
30. Yu, P., Merrick, L., Nuti, G., Campos, D.: Arctic-Embed 2.0: Multilingual Retrieval Without Compromise (2024). https://arxiv.org/abs/2412.04506 [↗](https://arxiv.org/abs/2412.04506)
31. Nussbaum, Z., Duderstadt, B.: Training Sparse Mixture Of Experts Text Embedding Models (2025). https://arxiv.org/abs/2502.07972 [↗](https://arxiv.org/abs/2502.07972)
32. Johnson, J., Douze, M., Jégou, H.: Billion-scale similarity search with GPUs. IEEE Transactions on Big Data 7 (3), 535–547 (2019) [our card](https://md.research.elamur.ai/free/en/papers/johnson-2017-billion-scale-similarity-search-gpus/)
33. Li, C., Liu, Z., Xiao, S., Shao, Y.: Making Large Language Models A Better Foundation For Dense Retrieval (2023) [find](https://scholar.google.com/scholar?q=Li%2C+C%2E%2C+Liu%2C+Z%2E%2C+Xiao%2C+S%2E%2C+Shao%2C+Y%2E%3A+Making+Large+Language+Models+A+Better+Foundation+For+Dense+Retrieval+%282023%29)
34. Landis, J.R., Koch, G.G.: The measurement of observer agreement for categorical data. Biometrics 33 (1), 159–174 (1977) [find](https://scholar.google.com/scholar?q=Landis%2C+J%2ER%2E%2C+Koch%2C+G%2EG%2E%3A+The+measurement+of+observer+agreement+for+categorical+data%2E+Biometrics+33+%281%29%2C+159%E2%80%93174+%281977%29)
35. Li, T., Zhang, G., Do, Q.D., Yue, X., Chen, W.: Long-context LLMs Struggle with Long In-context Learning (2024). https://arxiv.org/abs/2404.02060 [↗](https://arxiv.org/abs/2404.02060) [our card](https://md.research.elamur.ai/free/en/papers/li-2024-long-context-llms-struggle-long/)
36. Jaroslawicz, D., Whiting, B., Shah, P., Maamari, K.: How Many Instructions Can LLMs Follow at Once? (2025). https://arxiv.org/abs/2507.11538 [↗](https://arxiv.org/abs/2507.11538)
37. Darwish, K., Habash, N., Abbas, M., Al-Khalifa, H., Al-Natsheh, H.T., Bouamor, H., Bouzoubaa, K., Cavalli-Sforza, V., El-Beltagy, S.R., El-Hajj, W., Jarrar, M., Mubarak, H.: A panoramic survey of natural language processing in the arab world. Commun. ACM 64 (4), 72–81 (2021) https://doi.org/10.1145/3447735 [↗](https://doi.org/10.1145/3447735)
38. Alayba, A.M.: Arabic natural language processing (nlp): A comprehensive review of challenges, techniques, and emerging trends. Computers 14 (11) (2025) https://doi.org/10.3390/computers14110497 [↗](https://doi.org/10.3390/computers14110497)