---
title: "Halluscoring 2026 — paper by Alansari et al. (2026)"
description: "HalluScoring 2026 is a shared task on whether systems can tell when an LLM's answer to an Arabic question contains fabricated or wrong… 32 references."
authors: ["Aisha Alansari", "Abdessalam Bouchekif", "Ahmed Hasanaath", "Salah Eddine Bekhouche", "Malak Alkhorasani", "Mohammed-En-Nadhir Zighem", "Saad Ezzini", "Hichem Telli", "Hend Al-Khalifa", "Muhammad Abdul-Mageed", "Hadid Abdenour", "Hamzah Luqman"]
year: 2026
venue: "arXiv:2609.38355"
source: "https://arxiv.org/abs/2609.38355"
doi: "10.48550/arxiv.2609.38355"
arxiv: "2609.38355"
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/alansari-2026-halluscoring-2026-first-shared-task/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/alansari-2026-halluscoring-2026-first-shared-task/)

[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/) / Halluscoring 2026: The first shared task on llms hallucination detection and answer verification

# Halluscoring 2026: The first shared task on llms hallucination detection and answer verification

**Authors:** [Aisha Alansari](https://md.research.elamur.ai/free/en/authors/aisha-alansari/), Abdessalam Bouchekif, [Ahmed Hasanaath](https://md.research.elamur.ai/free/en/authors/ahmed-hasanaath/), [Salah Eddine Bekhouche](https://md.research.elamur.ai/free/en/authors/salah-eddine-bekhouche/), Malak Alkhorasani, Mohammed-En-Nadhir Zighem, [Saad Ezzini](https://md.research.elamur.ai/free/en/authors/saad-ezzini/), Hichem Telli, [Hend Al-Khalifa](https://md.research.elamur.ai/free/en/authors/hend-al-khalifa/), [Muhammad Abdul-Mageed](https://md.research.elamur.ai/free/en/authors/muhammad-abdul-mageed/), Hadid Abdenour and [Hamzah Luqman](https://md.research.elamur.ai/free/en/authors/hamzah-luqman/) · **Year:** 2026 · **Venue:** arXiv:2609.38355 · **Licence:** [arXiv — non-exclusive license to distribute](http://arxiv.org/licenses/nonexclusive-distrib/1.0/)

**Links:** [Publisher's page](https://arxiv.org/abs/2609.38355) · [PDF](https://arxiv.org/pdf/2609.38355) · [DOI 10.48550/arxiv.2609.38355](https://doi.org/10.48550/arxiv.2609.38355) · [arXiv 2609.38355](https://arxiv.org/abs/2609.38355) · [OpenAlex](https://openalex.org/W7215058096)

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

## Summary
HalluScoring 2026 is a shared task on whether systems can tell when an LLM's answer to an Arabic question contains fabricated or wrong facts, and whether they can then recover the true answer. It has two tasks. Task 1 is binary hallucination detection in two settings. In the first, test questions are new. In the second, both the questions and the generating LLMs are new. Task 2 combines detection with choosing the correct answer from six closely related options. One subtask covers Islamic knowledge and the other covers general knowledge (history, science, geography). The data come from two expert-annotated Arabic resources, HalluScore and HalluTruthQA. Systems had to be fine-tuned open models of at most 13B parameters, with no retrieval allowed. Organizers ran the submitted notebooks on hidden test sets. Thirteen teams took part. Detection scores stayed moderate, with AUC-ROC around 0.77. Access to the candidate answers raised Task 2 scores for nearly every system. The overall lesson is that flagging an error and knowing the correct fact are separate abilities, and robustness to distribution shift is still an open problem.

## Key points
- Participation: 13 teams, 10 of which submitted system description papers. Four separate CodaBench competitions were hosted, one per subtask. Training and development data were released on 16 May 2026.
- Task 1 winners by AUC-ROC:
  - REGLAT, 0.7717 on unseen questions (Subtask 1.1).
  - NAMAA, 0.7670 on unseen LLMs plus unseen questions (Subtask 1.2).
- Task 2 winner: Anhnamxtanh, a QLoRA-tuned ALLaM-7B with a cross-check verification step, ranked first in both subtasks.
  - Subtask 2.1: 0.8824 assisted, 0.8593 blind.
  - Subtask 2.2: 0.8565 assisted, 0.8348 blind.
- Effect of seeing the candidates: excluding the baseline, mean assisted scores beat blind scores by about 5.99% (Islamic) and 15.12% (general). Hashtag AI was the only exception. Its Gemma 4 12B, trained on contrastive examples from an offline knowledge-based teacher, scored higher blind (0.8262) than assisted (0.8124).
- Rules: open base models fine-tuned and capped at 13B parameters; no RAG or search at inference; external fine-tuning data allowed only if disclosed.

## Interesting ideas
- Splitting data so that the generators seen in testing never appear in training shows whether a detector learned real error signals or just the habits of particular models.
- Scoring each system both with and without the answer options turns the gap between the two scores into a measure of how much the system leans on the options.
- Keeping "is this claim wrong?" apart from "which fact is right?" made systems more stable when candidates were withheld, as with DzairVerse's separated design.
- Organizers kept test sets hidden and executed the participants' notebooks themselves, so no test data could leak into development.
- Distractors were built to be near-misses of the correct answer (wrong date, entity, number), which makes option choice demand precise fact checking.

## For Mova

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

- **I1** (learn): Hold out one witness parser entirely when inducing repair rules from silver data, to check whether the rules fix genuine errors rather than one parser's quirks.
- **I2** (answer): Run the school-test readers once without the choices and once with them, and treat a large gap as a sign the answer is not really derived from the world of the text.
- **I3** (world): Check a claim against the text's world state as its own explainable step, before and independently of choosing an answer, so each failure can be located.
- **I4** (learn): Lock evaluation sets behind a harness the training and self-play loops cannot read, so that reported gains cannot be contaminated by those sets.
- **I5** (back): Produce hard distractors by perturbing single nodes of the meaning graph and regenerating text, giving controlled negative examples for evaluating the readers.

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

## Institutions

- [Hamad bin Khalifa University](https://md.research.elamur.ai/free/en/universities/hamad-bin-khalifa-university/) — University, Doha, Qatar · 71 works · 12,942 citations
- [University of the Basque Country](https://md.research.elamur.ai/free/en/universities/university-of-the-basque-country/) — University, Leioa, Spain · 40 works · 16,234 citations
- [University of British Columbia](https://md.research.elamur.ai/free/en/universities/university-of-british-columbia/) — University, Vancouver, Canada · 46 works · 33,673 citations
- [King Fahd University of Petroleum and Minerals](https://md.research.elamur.ai/free/en/universities/king-fahd-university-of-petroleum-and-minerals/) — University, Dhahran, Saudi Arabia · 25 works · 4,450 citations
- [King Saud University](https://md.research.elamur.ai/free/en/universities/king-saud-university/) — University, Riyadh, Saudi Arabia · 16 works · 293 citations
- [Imam Abdulrahman Bin Faisal University](https://md.research.elamur.ai/free/en/universities/imam-abdulrahman-bin-faisal-university/) — University, Dammam, Saudi Arabia · 3 works · 40 citations
- [Universiti Malaysia Kelantan](https://md.research.elamur.ai/free/en/universities/universiti-malaysia-kelantan/) — University, Kota Bharu, Malaysia · 2 works · 1 citation
- [University of Biskra](https://md.research.elamur.ai/free/en/universities/university-of-biskra/) — University, Biskra, Algeria · 2 works · 1 citation

📚 Shelves (6): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 9,086 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 8,889 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 9,296 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 9,206 works · [🤝 Social Sciences](https://md.research.elamur.ai/free/en/shelves/social-sciences/) · 744 works · [🤝 Sociology and Political Science](https://md.research.elamur.ai/free/en/shelves/social-sciences/sociology-and-political-science/) · 221 works

🧭 Topics (1): [Semitic languages](https://md.research.elamur.ai/free/en/topics/semitic-languages/) · 1,484 works

## Cite

Cite as: Aisha Alansari, Abdessalam Bouchekif, Ahmed Hasanaath, Salah Eddine Bekhouche, Malak Alkhorasani, Mohammed-En-Nadhir Zighem, Saad Ezzini, Hichem Telli, Hend Al-Khalifa, Muhammad Abdul-Mageed, Hadid Abdenour and Hamzah Luqman. 2026. Halluscoring 2026: The first shared task on llms hallucination detection and answer verification. arXiv:2609.38355.

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

```bibtex
@misc{alansari2026halluscoring,
  title = {{Halluscoring 2026: The first shared task on llms hallucination detection and answer verification}},
  author = {Aisha Alansari, Abdessalam Bouchekif, Ahmed Hasanaath, Salah Eddine Bekhouche, Malak Alkhorasani, Mohammed-En-Nadhir Zighem, Saad Ezzini, Hichem Telli, Hend Al-Khalifa, Muhammad Abdul-Mageed, Hadid Abdenour and Hamzah Luqman},
  year = {2026},
  doi = {10.48550/arxiv.2609.38355},
  eprint = {2609.38355},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2609.38355}
}
```

- [← previous 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/)
- [next → Harder than Finding the Lost Sheep? Towards Automatically Suggesting Deliberate Metaphor Annotations in German Sermons](https://md.research.elamur.ai/free/en/papers/quante-2026-harder-than-finding-lost-sheep/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/7/)

## References

32 references, 7 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Samir Abdaljalil, Erchin Serpedin, Khalid Qaraqe, and Hasan Kurban. 2025. Evaluating multilingual and code-switched alignment in llms via synthetic natural language inference. arXiv preprint arXiv:2508.14735. [↗](https://arxiv.org/abs/2508.14735)
2. Firoj Alam, Shammur Absar Chowdhury, Sabri Boughorbel, and Maram Hasanain. 2024. Llms for low resource languages in multilingual, multimodal and dialectal settings. In Proceedings of the 18th conference of the European chapter of the association for computational linguistics: tutorial abstracts, pages 27–33. [find](https://scholar.google.com/scholar?q=Firoj+Alam%2C+Shammur+Absar+Chowdhury%2C+Sabri+Boughorbel%2C+and+Maram+Hasanain%2E+2024%2E+Llms+for+low+resource+languages+in+multilingual%2C+multimodal+and+dialectal+settings%2E+In+Proceedings+of+the+18th+conferen)
3. Aisha Alansari, Malak Alkhorasani, and Hamzah Luqman. 2026. Crosshallu: Do hallucination signals generalize across languages and domains in large language model’s internals? arXiv preprint arXiv:2607.04029. [↗](https://arxiv.org/abs/2607.04029)
4. Aisha Alansari and Hamzah Luqman. 2025. Arahallueval: A fine-grained hallucination evaluation framework for arabic llms. In Proceedings of The Third Arabic Natural Language Processing Conference, pages 148–161. [↗](https://doi.org/10.48448/hkxc-4z70)
5. Aisha Alansari and Hamzah Luqman. 2026a. Halluscore: Large language model hallucination question answering benchmark. arXiv preprint arXiv:2605.17007. [↗](https://arxiv.org/abs/2605.17007) [our card](https://md.research.elamur.ai/free/en/papers/alansari-2026-halluscore-large-language-model-hallucination/)
6. Aisha Alansari and Hamzah Luqman. 2026b. Large language models hallucination: A comprehensive survey. Computer Science Review, 61:100970. [↗](https://doi.org/10.1016/j.cosrev.2026.100970)
7. Najd Albalhy, Hassan Barmandah, Karim Elsayed, Hewida Ali, Yousef Ammar Abdullah, and Omer Nacar. 2026. NAMAA Community at HalluScoring-2026: Beyond predictive performance: Reliability and generalization in Arabic hallucination detection. In Proceedings of The Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Najd+Albalhy%2C+Hassan+Barmandah%2C+Karim+Elsayed%2C+Hewida+Ali%2C+Yousef+Ammar+Abdullah%2C+and+Omer+Nacar%2E+2026%2E+NAMAA+Community+at+HalluScoring%2D2026%3A+Beyond+predictive+performance%3A+Reliability+and+generalizat)
8. Ahmed Alzubaidi, Shaikha Alsuwaidi, Basma El Amel Boussaha, Leen AlQadi, Omar Alkaabi, Mohammed Alyafeai, Hamza Alobeidli, and Hakim Hacid. 2025. Evaluating arabic large language models: A survey of benchmarks, methods, and gaps. arXiv preprint arXiv:2510.13430. [↗](https://arxiv.org/abs/2510.13430)
9. Nsrin Ashraf, Mariam Mohamed Mahmoud, Mariam Francies, Ahmed M. Fetouh, and Hamada Nayel. 2026. REGLAT at HalluScoring-2026: A multistage approach for Arabic hallucination detection. In Proceedings of The Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Nsrin+Ashraf%2C+Mariam+Mohamed+Mahmoud%2C+Mariam+Francies%2C+Ahmed+M%2E+Fetouh%2C+and+Hamada+Nayel%2E+2026%2E+REGLAT+at+HalluScoring%2D2026%3A+A+multistage+approach+for+Arabic+hallucination+detection%2E+In+Proceedings+of)
10. Salah Eddine Bekhouche, Abdessalam Bouchekif, Hichem Telli, Mohammed-En-Nadhir Zighem, and Abdenour Hadid. 2026. Hallutruthqa-4k: A fine-grained corpus and annotation process for arabic hallucination detection and truth verification. arXiv preprint arXiv:2608.03966. [↗](https://arxiv.org/abs/2608.03966) [our card](https://md.research.elamur.ai/free/en/papers/bekhouche-2026-hallutruthqa-4k-fine-grained-corpus/)
11. Abdessalam Bouchekif, Samer Rashwani, Heba Sbahi, Shahd Gaben, Mutaz Al Khatib, and Mohammed Ghaly. 2025. Assessing large language models on islamic legal reasoning: Evidence from inheritance law evaluation. In Proceedings of The Third Arabic Natural Language Processing Conference, pages 246– 257. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Abdessalam+Bouchekif%2C+Samer+Rashwani%2C+Heba+Sbahi%2C+Shahd+Gaben%2C+Mutaz+Al+Khatib%2C+and+Mohammed+Ghaly%2E+2025%2E+Assessing+large+language+models+on+islamic+legal+reasoning%3A+Evidence+from+inheritance+law+eval)
12. Abdessalam Bouchekif, Mohammed-En-Nadhir Zighem, Salah Eddine Bekhouche, Hichem Telli, Somaya Eltanbouly, Shahd Gaben, Heba Sbahi, Samer Rashwani, Mutaz Al-Khatib, Emad Mohamed, and 1 others. 2026. Hallutruthqa: A fine-grained benchmark for hallucination detection, localization, and explanation in arabic question answering. arXiv preprint arXiv:2607.20219. [↗](https://arxiv.org/abs/2607.20219)
13. Yazid Bounab, Azeddine Benlamoudi, Wassim Swaileh, and Mia Sivén. 2026. HelsinkiTeam at HalluScoring-2026: Domain-adaptive Arabic hallucination detection and answer selection with ALLaM. In Proceedings of The Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Yazid+Bounab%2C+Azeddine+Benlamoudi%2C+Wassim+Swaileh%2C+and+Mia+Siv%C3%A9n%2E+2026%2E+HelsinkiTeam+at+HalluScoring%2D2026%3A+Domain%2Dadaptive+Arabic+hallucination+detection+and+answer+selection+with+ALLaM%2E+In+Proceeding)
14. Anupam Chaudhuri, Anj Simmons, Satwik Mukherjee, and Rongrong Song. 2026. Hashtag AI at HalluScoring-2026: Distilling knowledge-graph retrieval into Gemma 4 for Arabic hallucination detection. In Proceedings of The Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Anupam+Chaudhuri%2C+Anj+Simmons%2C+Satwik+Mukherjee%2C+and+Rongrong+Song%2E+2026%2E+Hashtag+AI+at+HalluScoring%2D2026%3A+Distilling+knowledge%2Dgraph+retrieval+into+Gemma+4+for+Arabic+hallucination+detection%2E+In+Proc)
15. Abdelmalek Chenini and Hacene Belhadef. 2026. Dzair-Verse at HalluScoring-2026: Leak-free QLoRA fine-tuning for Arabic hallucination detection and truth selection. In Proceedings of The Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Abdelmalek+Chenini+and+Hacene+Belhadef%2E+2026%2E+Dzair%2DVerse+at+HalluScoring%2D2026%3A+Leak%2Dfree+QLoRA+fine%2Dtuning+for+Arabic+hallucination+detection+and+truth+selection%2E+In+Proceedings+of+The+Fourth+Arabic+)
16. Ahmed Abul Hasanaath, Aisha Alansari, Ahmed Ashraf, Salmane Chafik, Hamzah Luqman, and Saad Ezzini. 2025. Arareasoner: Evaluating reasoning-based llms for arabic nlp. In EMNLP (Findings), pages 18898– 18914. [find](https://scholar.google.com/scholar?q=Ahmed+Abul+Hasanaath%2C+Aisha+Alansari%2C+Ahmed+Ashraf%2C+Salmane+Chafik%2C+Hamzah+Luqman%2C+and+Saad+Ezzini%2E+2025%2E+Arareasoner%3A+Evaluating+reasoning%2Dbased+llms+for+arabic+nlp%2E+In+EMNLP+%28Findings%29%2C+pages+18898%E2%80%93)
17. Lei Huang, Weijiang Yu, Weitao Ma, Weihong Zhong, Zhangyin Feng, Haotian Wang, Qianglong Chen, Weihua Peng, Xiaocheng Feng, Bing Qin, and 1 others. 2025. A survey on hallucination in large language models: Principles, taxonomy, challenges, and open questions. ACM Transactions on Information Systems, 43(2):1–55. [our card](https://md.research.elamur.ai/free/en/papers/huang-2023-survey-hallucination-large-language-models/)
18. Ziwei Ji, Nayeon Lee, Rita Frieske, Tiezheng Yu, Dan Su, Yan Xu, Etsuko Ishii, Ye Jin Bang, Andrea Madotto, and Pascale Fung. 2023. Survey of hallucination in natural language generation. ACM computing surveys, 55(12):1–38. [↗](https://doi.org/10.1145/3571730) [our card](https://md.research.elamur.ai/free/en/papers/ji-2022-survey-hallucination-natural-language-generation/)
19. Faiz Kraimia and Aicha Eutamene. 2026. ShadowDZ at HalluScoring-2026: A fine-tuned Fanar approach for Islamic knowledge hallucination detection. In Proceedings of The Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Faiz+Kraimia+and+Aicha+Eutamene%2E+2026%2E+ShadowDZ+at+HalluScoring%2D2026%3A+A+fine%2Dtuned+Fanar+approach+for+Islamic+knowledge+hallucination+detection%2E+In+Proceedings+of+The+Fourth+Arabic+Natural+Language+Pr)
20. Kenneth Li, Oam Patel, Fernanda Viégas, Hanspeter Pfister, and Martin Wattenberg. 2023. Inference-time intervention: Eliciting truthful answers from a language model. Advances in neural information processing systems, 36:41451–41530. [our card](https://md.research.elamur.ai/free/en/papers/li-2023-inference-time-intervention-eliciting-truthful/)
21. Malak Mashabi, Shahad Al-Khalifa, and Hend Al-Khalifa. 2026. A survey of large language models for arabic language and its dialects. ACM Transactions on Asian and Low-Resource Language Information Processing, 25(7):1–47. [find](https://scholar.google.com/scholar?q=Malak+Mashabi%2C+Shahad+Al%2DKhalifa%2C+and+Hend+Al%2DKhalifa%2E+2026%2E+A+survey+of+large+language+models+for+arabic+language+and+its+dialects%2E+ACM+Transactions+on+Asian+and+Low%2DResource+Language+Information+Pro)
22. Joshua Maynez, Shashi Narayan, Bernd Bohnet, and Ryan McDonald. 2020. On faithfulness and factuality in abstractive summarization. In Proceedings of the 58th annual meeting of the association for computational linguistics, pages 1906–1919. [find](https://scholar.google.com/scholar?q=Joshua+Maynez%2C+Shashi+Narayan%2C+Bernd+Bohnet%2C+and+Ryan+McDonald%2E+2020%2E+On+faithfulness+and+factuality+in+abstractive+summarization%2E+In+Proceedings+of+the+58th+annual+meeting+of+the+association+for+comp)
23. Timothee Mickus, Elaine Zosa, Raúl Vázquez, Teemu Vahtola, Jörg Tiedemann, Vincent Segonne, Alessandro Raganato, and Marianna Apidianaki. 2024. Semeval-2024 task 6: Shroom, a shared-task on hallucinations and related observable overgeneration mistakes. In Proceedings of the 18th International Workshop on Semantic Evaluation (SemEval-2024), pages 1979–1993. [find](https://scholar.google.com/scholar?q=Timothee+Mickus%2C+Elaine+Zosa%2C+Ra%C3%BAl+V%C3%A1zquez%2C+Teemu+Vahtola%2C+J%C3%B6rg+Tiedemann%2C+Vincent+Segonne%2C+Alessandro+Raganato%2C+and+Marianna+Apidianaki%2E+2024%2E+Semeval%2D2024+task+6%3A+Shroom%2C+a+shared%2Dtask+on+hallucinat)
24. Hamdy Mubarak, Hend Al-Khalifa, and Khaloud Suliman Alkhalefah. 2024. Halwasa: Quantify and analyze hallucinations in large language models: Arabic as a case study. In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), pages 8008–8015. [find](https://scholar.google.com/scholar?q=Hamdy+Mubarak%2C+Hend+Al%2DKhalifa%2C+and+Khaloud+Suliman+Alkhalefah%2E+2024%2E+Halwasa%3A+Quantify+and+analyze+hallucinations+in+large+language+models%3A+Arabic+as+a+case+study%2E+In+Proceedings+of+the+2024+Joint+In)
25. Hamdy Mubarak, Rana Malhas, Watheq Mansour, Abubakr Mohamed, Mahmoud Fawzi, Majd Hawasly, Tamer Elsayed, Kareem Mohamed Darwish, and Walid Magdy. 2025. Islamiceval 2025: The first shared task of capturing llms hallucination in islamic content. In Proceedings of The Third Arabic Natural Language Processing Conference: Shared Tasks, pages 480–493. [find](https://scholar.google.com/scholar?q=Hamdy+Mubarak%2C+Rana+Malhas%2C+Watheq+Mansour%2C+Abubakr+Mohamed%2C+Mahmoud+Fawzi%2C+Majd+Hawasly%2C+Tamer+Elsayed%2C+Kareem+Mohamed+Darwish%2C+and+Walid+Magdy%2E+2025%2E+Islamiceval+2025%3A+The+first+shared+task+of+captu)
26. Tu Nam Nguyen and Hiroki Takahashi. 2026. Anhnamxtanh at HalluScoring-2026: Counterfactual cross-checking for hallucination detection and answer verification in Islamic knowledge. In Proceedings of The Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Tu+Nam+Nguyen+and+Hiroki+Takahashi%2E+2026%2E+Anhnamxtanh+at+HalluScoring%2D2026%3A+Counterfactual+cross%2Dchecking+for+hallucination+detection+and+answer+verification+in+Islamic+knowledge%2E+In+Proceedings+of+Th)
27. Anjali R and Anand Kumar Madasamy. 2026. SCaLAR@NITK at HalluScoring-2026: A reference-free generative approach to hallucination detection via teacher-student distillation. In Proceedings of The Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Anjali+R+and+Anand+Kumar+Madasamy%2E+2026%2E+SCaLAR%40NITK+at+HalluScoring%2D2026%3A+A+reference%2Dfree+generative+approach+to+hallucination+detection+via+teacher%2Dstudent+distillation%2E+In+Proceedings+of+The+Fourt)
28. Ibad-ur-Rehman Rashid. 2026. AyahVerse at HalluScoring-2026: Category-aware Arabic hallucination detection with multi-layer representations. In Proceedings of The Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Ibad%2Dur%2DRehman+Rashid%2E+2026%2E+AyahVerse+at+HalluScoring%2D2026%3A+Category%2Daware+Arabic+hallucination+detection+with+multi%2Dlayer+representations%2E+In+Proceedings+of+The+Fourth+Arabic+Natural+Language+Proces)
29. Abdellah Zakaria Sellam, Ahmed Marouane Djouama, Fadi Abdeladhim Zidi, Ilyes Benaissa, Mohamed Amine Kerkouri, Marouane Tliba, and Aladine Chetouani. 2026. Vescera at HalluScoring-2026: Privileged-information distillation with contrastive representation alignment for Arabic hallucination detection. In Proceedings of The Fourth Arabic Natural Language Processing Conference: Shared Tasks, Budapest, Hungary. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Abdellah+Zakaria+Sellam%2C+Ahmed+Marouane+Djouama%2C+Fadi+Abdeladhim+Zidi%2C+Ilyes+Benaissa%2C+Mohamed+Amine+Kerkouri%2C+Marouane+Tliba%2C+and+Aladine+Chetouani%2E+2026%2E+Vescera+at+HalluScoring%2D2026%3A+Privileged%2Dinf)
30. Fanar Team, Ummar Abbas, Mohammad Shahmeer Ahmad, Firoj Alam, Enes Altinisik, Ehsannedin Asgari, Yazan Boshmaf, Sabri Boughorbel, Sanjay Chawla, Shammur Chowdhury, Fahim Dalvi, Kareem Darwish, Nadir Durrani, Mohamed Elfeky, Ahmed Elmagarmid, Mohamed Eltabakh, Masoomali Fatehkia, Anastasios Fragkopoulos, Maram Hasanain, and 23 others. 2025. Fanar: An arabic-centric multimodal generative ai platform. [our card](https://md.research.elamur.ai/free/en/papers/abbas-2025-fanar-arabic-centric-multimodal-generative/)
31. Raúl Vázquez, Timothee Mickus, Elaine Zosa, Teemu Vahtola, Jörg Tiedemann, Aman Sinha, Vincent Segonne, Fernando Sánchez-Vega, Alessandro Raganato, Jindřich Libovickỳ, and 1 others. 2025. Semeval-2025 task 3: Mu-shroom, the multilingual shared task on hallucinations and related observable overgeneration mistakes. arXiv preprint arXiv:2504.11975. [↗](https://arxiv.org/abs/2504.11975)
32. Junchao Wu, Shu Yang, Runzhe Zhan, Yulin Yuan, Lidia Sam Chao, and Derek Fai Wong. 2025. A survey on llm-generated text detection: Necessity, methods, and future directions. Computational Linguistics, 51(1):275–338. [↗](https://doi.org/10.1162/coli_a_00549) [our card](https://md.research.elamur.ai/free/en/papers/wu-2025-survey-llm-generated-text-detection/)

*Updated 24 October 2026*