--- title: "LoRAD — paper by Zain (2026) — summary and ideas" description: "The paper describes the BUSTED team's entry in the AbjadGenEval Shared Task (ArabicNLP 2026). 0.8877 accuracy. 11 references." authors: ["Ali Zain"] year: 2026 venue: "Proceedings of the 2nd Workshop on NLP for Languages Using Arabic Script" source: "https://aclanthology.org/2026.abjadnlp-1.57/" doi: "10.18653/v1/2026.abjadnlp-1.57" licence: "CC BY 4.0 (ACL Anthology)" licence_url: "https://creativecommons.org/licenses/by/4.0/" summary_licence: "Elamur Cosmic License" canonical: "https://research.elamur.ai/free/en/papers/zain-2026-lorad-low-resource-ai-generated/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/zain-2026-lorad-low-resource-ai-generated/) [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/14/) / LoRAD: Low-Resource AI-Generated Text Detection with XLM-RoBERTa # LoRAD: Low-Resource AI-Generated Text Detection with XLM-RoBERTa **Authors:** [Ali Zain](https://md.research.elamur.ai/free/en/authors/ali-zain/) · **Year:** 2026 · **Venue:** [Proceedings of the 2nd Workshop on NLP for Languages Using Arabic Script](https://md.research.elamur.ai/free/en/venues/abjadnlp/) ([Workshop on NLP for Languages Using Arabic Script 2026](https://md.research.elamur.ai/free/en/venues/abjadnlp/2026/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/) **Links:** [Publisher's page](https://aclanthology.org/2026.abjadnlp-1.57/) · [PDF](https://aclanthology.org/2026.abjadnlp-1.57.pdf) · [DOI 10.18653/v1/2026.abjadnlp-1.57](https://doi.org/10.18653/v1/2026.abjadnlp-1.57) · [OpenAlex](https://openalex.org/W7140084358) **Citations:** 1 citations (OpenAlex, 2026-10-14) ## Summary The paper describes the BUSTED team's entry in the AbjadGenEval Shared Task (ArabicNLP 2026). The task asks systems to decide whether an Arabic or Urdu news or literary text was written by a person or by a language model. The author does not design a new architecture. Instead, one multilingual backbone, XLM-RoBERTa with a standard classification head, is fine-tuned separately for each language, and only the data handling changes. For Arabic, the official training set is expanded with an organizer-provided Hugging Face corpus of machine-generated abstracts, and inputs are capped at 256 tokens. For Urdu, only the official data is used. The text is kept largely uncleaned, on the assumption that generator artifacts remain useful signals, and inputs can reach 512 tokens. The system ranked first on Urdu (F1 0.8877) and third on Arabic (F1 0.8866). The work shows that a carefully configured off-the-shelf multilingual encoder can be a competitive detector for low-resource Arabic-script languages. However, the paper provides no baselines or ablations to confirm the effect of its preprocessing choices. ## Key points - **Data:** The Arabic training set has 5,278 balanced samples. An external set (KFUPM-JRCAI/arabic-generated-abstracts) adds 12,870 samples, of which 2,574 are human and 10,296 are machine. The Arabic test set has 200 samples. Urdu has 11,904 balanced training samples and 2,630 test samples. - **Training setup:** Learning rate 2e-5, AdamW, weight decay 0.01, up to 6 epochs, and batch size 4–16 depending on memory. Training ran on Kaggle with dual NVIDIA T4 GPUs. - **Urdu results (1st place):** F1 0.8877, accuracy 0.8878, precision 0.8834, recall 0.8922. - **Arabic results (3rd place):** F1 0.8866, accuracy 0.8850, precision 0.8737, recall 0.9000. - **Limitations:** The paper tests a single model with no baseline comparisons. Evaluation covers only the shared-task data, and the code had not yet been released at publication. *Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.* 📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 16,039 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 15,891 works 🧭 Topics (2): [Semitic languages](https://md.research.elamur.ai/free/en/topics/semitic-languages/) · 2,908 works · [🌱 Low-resource languages](https://md.research.elamur.ai/free/en/topics/low-resource-languages/) · 267 works ## Cite Cite as: Ali Zain. 2026. LoRAD: Low-Resource AI-Generated Text Detection with XLM-RoBERTa. Proceedings of the 2nd Workshop on NLP for Languages Using Arabic Script. The paper itself, not this page. BibTeX from the library's record of the work. ```bibtex @inproceedings{zain2026lorad, title = {{LoRAD: Low-Resource AI-Generated Text Detection with XLM-RoBERTa}}, author = {Ali Zain}, booktitle = {Proceedings of the 2nd Workshop on NLP for Languages Using Arabic Script}, year = {2026}, doi = {10.18653/v1/2026.abjadnlp-1.57}, url = {https://doi.org/10.18653/v1/2026.abjadnlp-1.57} } ``` - [← previous looploop at SemEval-2026 Task 3: A Dimensional Aspect-Based Sentiment System with DeBERTa Regression and Qwen3 Instruction Fine-Tuning](https://md.research.elamur.ai/free/en/papers/yang-2026-looploop-semeval-2026-task-3/) - [next → LTRC-IIIT at MEDIQA-SYNUR 2026: Benchmarking a Fully Local, Training-Free RAG Pipeline](https://md.research.elamur.ai/free/en/papers/vaish-2026-ltrc-iiit-mediqa-synur-2026/) - [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/14/) ## References 11 references, 7 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. Shadi Abudalfa, Saad Ezzini, Ahmed Abdelali, Hamza Alami, Abdessamad Benlahbib, Salmane Chafik, Mo El-Haj, Abdelkader El Mahdaouy, Mustafa Jarrar, Salima Lamsiyah, and Hamzah Luqman. 2025. [The AraGenEval shared task on Arabic authorship style transfer and AI generated text detection](https://md.research.elamur.ai/free/en/papers/abudalfa-2025-arageneval-shared-task-arabic-authorship/). In Proceedings of The Third Arabic Natural Language Processing Conference at EMNLP 2025, Suzhou, China. Association for Computational Linguistics. 2. Alexis Conneau, Kartikay Khandelwal, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke Zettlemoyer, and Veselin Stoyanov. 2020. Unsupervised cross-lingual representation learning at scale. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 8440– 8451. [find](https://scholar.google.com/scholar?q=Alexis+Conneau%2C+Kartikay+Khandelwal%2C+Naman+Goyal%2C+Vishrav+Chaudhary%2C+Guillaume+Wenzek%2C+Francisco+Guzm%C3%A1n%2C+Edouard+Grave%2C+Myle+Ott%2C+Luke+Zettlemoyer%2C+and+Veselin+Stoyanov%2E+2020%2E+Unsupervised+cross%2Dlingu) 3. Saad Ezzini, Irfan Ahmed, Salmane Chafik, Shadi Abudalfa, Mo El-Haj, Ahmed Abdelali, Mustafa Jarrar, Nadir Durrani, Hassan Sajjad, and Farah Adeeba. 2026. [Abjadgeneval: Abjad ai generated text detection shared task for languages using arabic script at abjadnlp 2026](https://md.research.elamur.ai/free/en/papers/ezzini-2026-abjadgeneval-abjad-ai-generated-text/). In Proceedings of the 2nd Workshop on NLP for Languages Using Arabic Script (AbjadNLP 2026), co-located with the 19th Conference of the European Chapter of the Association for Computational Linguistics (EACL 2026), Rabat, Morocco. visualization of generated text. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics: System Demonstrations, pages 111–116. 4. Daphne Ippolito, Daniel Duckworth, Chris Callison-Burch, and Douglas Eck. 2020. [Automatic detection of generated text is easiest when humans are fooled](https://md.research.elamur.ai/free/en/papers/ippolito-2019-automatic-detection-generated-text-easiest/). In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 1808– 1822. 5. Salima Lamsiyah, Saad Ezzini, Abdelkader El Mahdaouy, Hamza Alami, Abdessamad Benlahbib, Samir El amrany, Salmane Chafik, and Hicham Hammouchi. 2025. M-DAIGT: A Shared Task on Multi-Domain Detection of AI-Generated Text. In Proceedings of the Shared Task on Multi-Domain Detection of AI-Generated Text, pages 1–9, Varna, Bulgaria. IN-COMA Ltd., Shoumen, Bulgaria. [find](https://scholar.google.com/scholar?q=Salima+Lamsiyah%2C+Saad+Ezzini%2C+Abdelkader+El+Mahdaouy%2C+Hamza+Alami%2C+Abdessamad+Benlahbib%2C+Samir+El+amrany%2C+Salmane+Chafik%2C+and+Hicham+Hammouchi%2E+2025%2E+M%2DDAIGT%3A+A+Shared+Task+on+Multi%2DDomain+Detection+o) 6. Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, and Veselin Stoyanov. 2019. [Roberta: A robustly optimized bert pretraining approach](https://md.research.elamur.ai/free/en/papers/liu-2019-roberta-robustly-optimized-bert-pretraining/). arXiv preprint arXiv:1907.11692. [↗](https://arxiv.org/abs/1907.11692) 7. Eric Mitchell, Yoonho Lee, Alexander Khazatsky, Christopher D. Manning, and Chelsea Finn. 2023. [Detectgpt: Zero-shot machine-generated text detection using probability curvature](https://md.research.elamur.ai/free/en/papers/mitchell-2023-detectgpt-zero-shot-machine-generated/). arXiv preprint arXiv:2301.11305. [↗](https://arxiv.org/abs/2301.11305) 8. Ali Zain, Sareem Farooqui, and Muhammad Rafi. 2025. [Busted at arageneval shared task: A comparative study of transformer-based models for arabic ai-generated text detection](https://md.research.elamur.ai/free/en/papers/zain-2025-busted-arageneval-shared-task-comparative/). In Proceedings of The Third Arabic Natural Language Processing Conference: Shared Tasks, pages 72–76. Accessed January 2026. [↗](https://doi.org/10.48550/arxiv.2510.20610) 9. Rowan Zellers, Ari Holtzman, Hannah Rashkin, Yonatan Bisk, Ali Farhadi, Franziska Roesner, and Yejin Choi. 2019. [Defending against neural fake news](https://md.research.elamur.ai/free/en/papers/zellers-2019-defending-against-neural-fake-news/). In Advances in Neural Information Processing Systems. [↗](https://doi.org/10.48550/arxiv.1905.12616) 10. Sareem Farooqui, Ali Zain, and Muhammad Rafi. 2025. Shared task on multi-domain detection of ai-generated text (m-daigt). In Proceedings of the Shared Task on Multi-Domain Detection of AI-Generated Text, pages 15–19. Accessed January 2026. [find](https://scholar.google.com/scholar?q=Sareem+Farooqui%2C+Ali+Zain%2C+and+Muhammad+Rafi%2E+2025%2E+Shared+task+on+multi%2Ddomain+detection+of+ai%2Dgenerated+text+%28m%2Ddaigt%29%2E+In+Proceedings+of+the+Shared+Task+on+Multi%2DDomain+Detection+of+AI%2DGenerated+Te) 11. Sebastian Gehrmann, Hendrik Strobelt, and Alexander M. Rush. 2019. Gltr: Statistical detection and [find](https://scholar.google.com/scholar?q=Sebastian+Gehrmann%2C+Hendrik+Strobelt%2C+and+Alexander+M%2E+Rush%2E+2019%2E+Gltr%3A+Statistical+detection+and) *Updated 1 November 2026*