Elamur Research⌕
☰
home / papers / 2026 / DetectRL-X: Towards Reliable Multilingual and Real-World LLM-Generated Text Detection

DetectRL-X: Towards Reliable Multilingual and Real-World LLM-Generated Text Detection

Paper · 2026 · Junchao Wu, Yefeng Liu, Chenyu Zhu et al.

Summary

Most benchmarks for detecting LLM-generated text (LGT) are English-only or cover languages unevenly. They also tend to ignore how people actually use LLMs, for example by asking a model to rework a draft. DetectRL-X addresses these gaps with 3.46 million samples in 8 languages from five language families and 6 misuse-prone domains. The texts come from 4 commercial generators. They are put through 8 multilingual paraphrasing and noise attacks, and they vary in length and in the kind of LLM refinement applied. The authors add a Ternary task that treats human text refined by an LLM (HLT) as its own class, separate from human-written text (HWT) and LGT. They test 12 detectors along 8 evaluation dimensions. Fine-tuned neural classifiers, especially mDeBERTa-Classifier, clearly outperform statistical detectors in the Binary task. However, they lose more accuracy across languages and when HLT is introduced. Real-world perturbations hurt both detector families, and most statistical methods fail at the Ternary task. The benchmark is a useful testbed for detectors that are more robust and treat languages more fairly.

Key points

  • Coverage:
    • Languages: English, German, Spanish, French, Portuguese, Russian, Arabic and Chinese, grouped into low, medium and high complexity.
    • Domains: Academic, News, Novel, Wiki, SEO and Webtext.
    • Generators: Deepseek-V3, Gemini-2.5-flash, GPT-4o and Qwen-Max.
  • Refinement operations and attacks:
    • Polishing, Expanding and Condensing are applied to both HWT and LGT.
    • Attacks include encoder, seq2seq and decoder paraphrasing, back translation, and character-level or zero-width perturbations.
  • Effect of text length: As texts get longer, mDeBERTa-Classifier's F1 rises from 96% to nearly 100% on the Binary task and from 66% to 97% on the Ternary task. On short texts, complex languages score only 45–55%.
  • Effect of refinement type: Expanding is the most damaging operation, lowering scores by 15–25% compared with the others. Binoculars reaches about 60% on Arabic and Chinese but about 75% on English and German.
  • Availability: Code and data are released in the AIDC-AI/Marco-LLM GitHub repository, for non-commercial academic use only.

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, topics and concepts

📚 Shelves 5The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 worksComputer Vision and Pattern Recognition · 3,486 works

🧭 Topics 1Research subjects and languages the work is about.Multilingual & cross-lingual NLP · 882 works

Cite

Cite as Junchao Wu, Yefeng Liu, Chenyu Zhu, Hao Zhang, Zeyu Wu, Tianqi Shi, Yichao Du, Longyue Wang, Weihua Luo, Jinsong Su and Derek F. Wong. 2026. DetectRL-X: Towards Reliable Multilingual and Real-World LLM-Generated Text Detection. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.acl-long.1773

@inproceedings{wu2026detectrl,
  title = {{DetectRL-X: Towards Reliable Multilingual and Real-World LLM-Generated Text Detection}},
  author = {Junchao Wu, Yefeng Liu, Chenyu Zhu, Hao Zhang, Zeyu Wu, Tianqi Shi, Yichao Du, Longyue Wang, Weihua Luo, Jinsong Su and Derek F. Wong},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.1773},
  url = {https://doi.org/10.18653/v1/2026.acl-long.1773}
}

References

45 references, 22 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Catherine Arnett and Benjamin Bergen. 2025. Why do language models perform worse for morphologically complex languages? In Proceedings of the 31st International Conference on Computational Linguistics, COLING 2025, Abu Dhabi, UAE, January 19-24, 2025, pages 6607–6623. Association for Computational L…
  2. Guangsheng Bao, Yanbin Zhao, Zhiyang Teng, Linyi Yang, and Yue Zhang. 2024. Fast-detectgpt: Efficient zero-shot detection of machine-generated text via conditional probability curvature. In The Twelfth International Conference on Learning Representations, ICLR 2024, Vienna, Austria, May 7-11, 2024.…
  3. Can Chen and Jun-Kun Wang. 2025. Online detection of llm-generated texts via sequential hypothesis testing by betting. In Forty-second International Conference on Machine Learning, ICML 2025, Vancouver, BC, Canada, July 13-19, 2025, Proceedings of Machine Learning Research. PMLR / OpenReview.net. find
  4. Guanhua Chen, Yutong Yao, Lidia S. Chao, Xuebo Liu, and Derek F. Wong. 2025a. SGIC: A self-guided iterative calibration framework for RAG. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), ACL 2025, Vienna, Austria, July 27 - August 1… find
  5. Jiaao Chen, Derek Tam, Colin Raffel, Mohit Bansal, and Diyi Yang. 2023. An empirical survey of data augmentation for limited data learning in NLP. Trans. Assoc. Comput. Linguistics, 11:191–211.
  6. Xin Chen, Junchao Wu, Shu Yang, Runzhe Zhan, Zeyu Wu, Ziyang Luo, Di Wang, Min Yang, Lidia S. Chao, and Derek F. Wong. 2025b. Repreguard: Detecting llm-generated text by revealing hidden representation patterns. CoRR, abs/2508.13152. find
  7. Andy Coenen, Luke Davis, Daphne Ippolito, Emily Reif, and Ann Yuan. 2021. Wordcraft: a human-ai collaborative editor for story writing. CoRR, abs/2107.07430. — [arXiv:2107.07430](https://arxiv.org/abs/2107.07430) find
  8. Gheorghe Comanici, Eric Bieber, Mike Schaekermann, Ice Pasupat, Noveen Sachdeva, Inderjit S. Dhillon, Marcel Blistein, Ori Ram, Dan Zhang, Evan Rosen, Luke Marris, Sam Petulla, Colin Gaffney, Asaf Aharoni, Nathan Lintz, Tiago Cardal Pais, Henrik Jacobsson, Idan Szpektor, Nan-Jiang Jiang, and 81 othe… — Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities
  9. DeepSeek-AI, Aixin Liu, Bei Feng, Bing Xue, Bingxuan Wang, Bochao Wu, Chengda Lu, Chenggang Zhao, Chengqi Deng, Chenyu Zhang, Chong Ruan, Damai Dai, Daya Guo, Dejian Yang, Deli Chen, Dongjie Ji, Erhang Li, Fangyun Lin, Fucong Dai, and 80 others. 2024. Deepseek-v3 technical report. CoRR, abs/2412.194… find
  10. Liam Dugan, Alyssa Hwang, Filip Trhlík, Andrew Zhu, Josh Magnus Ludan, Hainiu Xu, Daphne Ippolito, and Chris Callison-Burch. 2024. RAID: A shared benchmark for robust evaluation of machine-generated text detectors. In Proceedings of the 62nd Annual Meeting of the Association for Computational Lingui…
  11. Lester Faigley and Stephen Witte. 1981. Analyzing revision. College Composition & Communication, 32(4):400–414. find
  12. Rudolph Flesch. 1948. A new readability yardstick. Journal of applied psychology, 32(3):221. find
  13. Sebastian Gehrmann, Hendrik Strobelt, and Alexander M. Rush. 2019. GLTR: statistical detection and visualization of generated text. In Proceedings of the 57th Conference of the Association for Computational Linguistics, ACL 2019, Florence, Italy, July 28 - August 2, 2019, Volume 3: System Demonstrat…
  14. Biyang Guo, Xin Zhang, Ziyuan Wang, Minqi Jiang, Jinran Nie, Yuxuan Ding, Jianwei Yue, and Yupeng Wu. 2023. How close is chatgpt to human experts? comparison corpus, evaluation, and detection. CoRR, abs/2301.07597.
  15. Hanxi Guo, Siyuan Cheng, Xiaolong Jin, Zhuo Zhang, Kaiyuan Zhang, Guanhong Tao, Guangyu Shen, and Xiangyu Zhang. 2024. Biscope: Ai-generated text detection by checking memorization of preceding tokens. In Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information P… find
  16. Abhimanyu Hans, Avi Schwarzschild, Valeriia Cherepanova, Hamid Kazemi, Aniruddha Saha, Micah Goldblum, Jonas Geiping, and Tom Goldstein. 2024. Spotting llms with binoculars: Zero-shot detection of machine-generated text. In Forty-first International Conference on Machine Learning, ICML 2024, Vienna,…
  17. Pengcheng He, Xiaodong Liu, Jianfeng Gao, and Weizhu Chen. 2021. Deberta: decoding-enhanced bert with disentangled attention. In 9th International Conference on Learning Representations, ICLR 2021, Virtual Event, Austria, May 3-7, 2021. OpenReview.net.
  18. Xinlei He, Xinyue Shen, Zeyuan Chen, Michael Backes, and Yang Zhang. 2024. Mgtbench: Benchmarking machine-generated text detection. In Proceedings of the 2024 on ACM SIGSAC Conference on Computer and Communications Security, CCS 2024, Salt Lake City, UT, USA, October 14-18, 2024, pages 2251–2265. AC… find
  19. Xinke Jiang, Yue Fang, Rihong Qiu, Haoyu Zhang, Yongxin Xu, Hao Chen, Wentao Zhang, Ruizhe Zhang, Yuchen Fang, Xu Chu, and 1 others. 2025. Tc-rag: Turing-complete rag’s case study on medical llm systems.
  20. Kalpesh Krishna, Yixiao Song, Marzena Karpinska, John Wieting, and Mohit Iyyer. 2023. Paraphrasing evades detectors of ai-generated text, but retrieval is an effective defense. In Advances in Neural Information Processing Systems 36: Annual Conference on Neural Information Processing Systems 2023, N…
  21. Tatsuki Kuribayashi, Takumi Ito, Jun Suzuki, and Kentaro Inui. 2020. Language models as an alternative evaluator of word order hypotheses: A case study in japanese. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, ACL 2020, Online, July 5-10, 2020, pages 48… find
  22. Jooyoung Lee, Thai Le, Jinghui Chen, and Dongwon Lee. 2023. Do language models plagiarize? In Proceedings of the ACM Web Conference 2023, WWW 2023, Austin, TX, USA, 30 April 2023 - 4 May 2023, pages 3637–3647. ACM. find
  23. Yafu Li, Qintong Li, Leyang Cui, Wei Bi, Zhilin Wang, Longyue Wang, Linyi Yang, Shuming Shi, and Yue Zhang. 2024. MAGE: machine-generated text detection in the wild. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), ACL 2024, Bangkok,…
  24. Yuqing Liang, Jiancheng Xiao, Wensheng Gan, and Philip S Yu. 2026. Watermarking techniques for large language models: A survey. Artificial Intelligence Review, 59(2):74. find
  25. 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. CoRR, abs/1907.11692.
  26. Dominik Macko, Róbert Móro, Adaku Uchendu, Jason Samuel Lucas, Michiharu Yamashita, Matús Pikuliak, Ivan Srba, Thai Le, Dongwon Lee, Jakub Simko, and Mária Bieliková. 2023. Multitude: Large-scale multilingual machine-generated text detection benchmark. In Proceedings of the 2023 Conference on Empiri… find
  27. OpenAI. 2023. GPT-4 technical report. abs/2303.08774.
  28. CoRR, OpenRouter. 2025. AI Model Rankings. find
  29. Gerrit Quaremba, Elizabeth Black, Denny Vrandecic, and Elena Simperl. 2025. Wetbench: A benchmark for detecting task-specific machine-generated text on wikipedia. CoRR, abs/2507.03373. of the Association for Computational Linguistics (Volume 1: Long Papers), ACL 2024, Bangkok, Thailand, August 11-16… find
  30. Yuxia Wang, Jonibek Mansurov, Petar Ivanov, Jinyan Su, Artem Shelmanov, Akim Tsvigun, Chenxi Whitehouse, Osama Mohammed Afzal, Tarek Mahmoud, Toru Sasaki, Thomas Arnold, Alham Fikri Aji, Nizar Habash, Iryna Gurevych, and Preslav Nakov. 2024b. M4: multi-generator, multi-domain, and multi-lingual blac…
  31. Rico Sennrich, Barry Haddow, and Alexandra Birch. 2016. Improving neural machine translation models with monolingual data. In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics, ACL 2016, August 7-12, 2016, Berlin, Germany, Volume 1: Long Papers. The Association… find
  32. Jason W. Wei and Kai Zou. 2019. EDA: easy data augmentation techniques for boosting performance on text classification tasks. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing, EMNLP-IJC…
  33. Irene Solaiman, Miles Brundage, Jack Clark, Amanda Askell, Ariel Herbert-Voss, Jeff Wu, Alec Radford, and Jasmine Wang. 2019. Release strategies and the social impacts of language models. CoRR, abs/1908.09203.
  34. Junchao Wu, Shu Yang, Runzhe Zhan, Yulin Yuan, Lidia S. Chao, and Derek Fai Wong. 2025a. A survey on llm-generated text detection: Necessity, methods, and future directions. Comput. Linguistics, 51(1):275–338.
  35. Jinyan Su, Terry Yue Zhuo, Di Wang, and Preslav Nakov. 2023. Detectllm: Leveraging log rank information for zero-shot detection of machine-generated text. In Findings of the Association for Computational Linguistics: EMNLP 2023, Singapore, December 6-10, 2023, pages 12395–12412. Association for Comp…
  36. Junchao Wu, Runzhe Zhan, Qianli Wang, Yulin Yuan, Lidia S. Chao, and Derek F. Wong. 2025b. Overview of the NLPCC 2025 shared task 1: Llm-generated text detection. In Natural Language Processing and Chinese Computing - 14th National CCF Conference, NLPCC 2025, Urumqi, China, August 7-9, 2025, Proceed… find
  37. Adaku Uchendu, Zeyu Ma, Thai Le, Rui Zhang, and Dongwon Lee. 2021. TURINGBENCH: A benchmark environment for turing test in the age of neural text generation. In Findings of the Association for Computational Linguistics: EMNLP 2021, Virtual Event / Punta Cana, Dominican Republic, 16-20 November, 2021…
  38. Saranya Venkatraman, Nafis Irtiza Tripto, and Dongwon Lee. 2025. Collabstory: Multi-llm collaborative story generation and authorship analysis. In Findings of the Association for Computational Linguistics: NAACL 2025, Albuquerque, New Mexico, USA, April 29 -May 4, 2025, Findings of ACL, pages 3665–3… find
  39. Yichen Wang, Shangbin Feng, Abe Bohan Hou, Xiao Pu, Chao Shen, Xiaoming Liu, Yulia Tsvetkov, and Tianxing He. 2024a. Stumbling blocks: Stress testing the robustness of machine-generated text detectors under attacks. In Proceedings of the 62nd Annual Meeting Junchao Wu, Runzhe Zhan, Derek F. Wong, Sh… find
  40. Junchao Wu, Runzhe Zhan, Derek F. Wong, Shu Yang, Xinyi Yang, Yulin Yuan, and Lidia S. Chao. 2024. Detectrl: Benchmarking llm-generated text detection in real-world scenarios. In Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Processing Systems 2024, Ne… find
  41. Yihuai Xu, Yongwei Wang, Yifei Bi, Huangsen Cao, Zhouhan Lin, Yu Zhao, and Fei Wu. 2025. Training-free llm-generated text detection by mining token probability sequences. In The Thirteenth International Conference on Learning Representations, ICLR 2025, Singapore, April 24-28, 2025. OpenReview.net. find
  42. An Yang, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chengyuan Li, Dayiheng Liu, Fei Huang, Haoran Wei, Huan Lin, Jian Yang, Jianhong Tu, Jianwei Zhang, Jianxin Yang, Jiaxi Yang, Jingren Zhou, Junyang Lin, Kai Dang, and 22 others. 2024. Qwen2.5 technical report. CoRR, abs/2412.1511… find
  43. Shu Yang, Jingyu Hu, Tong Li, Hanqi Yan, Wenxuan Wang, and Di Wang. 2026. Automonitor-bench: Evaluating the reliability of llm-based misbehavior monitor. CoRR, abs/2601.05752.
  44. Jiayi Zhang, Shu Yang, Junchao Wu, Derek F. Wong, and Di Wang. 2025. Understanding and mitigating political stance cross-topic generalization in large language models. Preprint, arXiv:2508.02360.
  45. Qihui Zhang, Chujie Gao, Dongping Chen, Yue Huang, Yixin Huang, Zhenyang Sun, Shilin Zhang, Weiye Li, Zhengyan Fu, Yao Wan, and Lichao Sun. 2024. Llm-as-a-coauthor: Can mixed human-written and machine-generated text be detected? In Findings of the Association for Computational Linguistics: NAACL 202… find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/wu-2026-detectrl-x-reliable-multilingual-real/