RusHallu-RAG: a Comprehensive Approach to Hallucination Detection in Russian-Language RAG Systems
Main Article Content
Abstract
This paper introduces RusHallu-RAG, the first comprehensive benchmark for detecting hallucinations in Russian-language RAG systems. The dataset of 1,000 query-answer pairs covers general knowledge and scientific domains with controlled context relevance balancing. A fine-grained taxonomy of six hallucination types is proposed and used for annotation at both response and span levels with human experts and LLMs.
For training a hallucination detector, a balanced dataset of 3,000 examples is collected using synthetic generation methods. A detector based on YandexGPT-5-Lite-8B is trained via SFT. We evaluate 15 open-weight and proprietary models alongside the trained detector. Results show that model size does not guarantee performance: medium-sized models (20–33B) sometimes outperform larger ones. The trained detector achieves an average accuracy of 63.7%, ranking second after proprietary Gemini-2.5-Pro, with a gain of 47.2 p.p. over the baseline. The gap between proprietary and open-weight models is reduced from 20–35 to 8.4 p.p., confirming the effectiveness of the approach. The benchmark, code, and trained model are publicly released.
Article Details
References
2. Agarwal V., et al. MedHalu: Hallucinations in Responses to Healthcare Queries by Large Language Models // Findings of the Association for Computational Linguistics: ACL 2025. 2025. https://doi.org/10.1609/icwsm.v20i1.42623
3. Liu T., Zhang Y., Brockett C., Mao Y., Sui Z., Chen W., Dolan B. A Token-Level Reference-Free Hallucination Detection Benchmark for Free-Form Text Generation // Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Vol. 1: Long Papers). Dublin, Ireland, May 2022. P. 6723–6737. https://doi.org/10.18653/v1/2022.acl-long.464
4. Li J., Cheng X., Zhao X., Nie J.-Y., Wen J.-R. HaluEval: A Large-Scale Hallucination Evaluation Benchmark for Large Language Models // Proc. of the 2023 Conference on Empirical Methods in Natural Language Processing. Singapore, December 2023. P. 6449–6464. https://doi.org/10.18653/v1/2023.emnlp-main.397
5. Niu C., Wu Y., Zhu J., Xu S., Shum K., Zhong R., Song J., Zhang T. RAGTruth: A Hallucination Corpus for Developing Trustworthy Retrieval-Augmented Language Models // Proc. of the 62nd Annual Meeting of the Association for Computational Linguistics (Vol. 1: Long Papers). Bangkok, Thailand, August 2024. P. 10862–10878. 5. https://doi.org/10.18653/v1/2024.acl-long.585
6. Zhu Z., Yang Y., Sun Z. HaluEval-Wild: Evaluating Hallucinations of Language Models in the Wild // Advances in Neural Information Processing Systems. 2024. Vol. 37. https://doi.org/10.48550/arXiv.2403.04307
7. Vazquez R., et al. SemEval-2025 Task 3: Mu-SHROOM, the Multilingual Shared-Task on Hallucinations and Related Observable Overgeneration Mistakes // Proc. of the 19th International Workshop on Semantic Evaluation (SemEval-2025). Vienna, Austria, July 2025. P. 2472–2497. URL: https://aclanthology.org/2025.semeval-1.322/.
8. Chernogorskii F., et al. DRAGOn: Designing RAG on periodically updated corpus // Proc. of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Vol. 4: Student Research Workshop). 2026. P. 622–638. https://doi.org/10.18653/v1/2026.eacl-srw.48
9. Xiong G., Jin Q., Lu Z., Zhang A. Benchmarking Retrieval-Augmented Generation for Medicine // Findings of the Association for Computational Linguistics: ACL 2024. Bangkok, Thailand, August 2024. С. 6233–6251. 9. https://doi.org/10.18653/v1/2024.findings-acl.372
10. Tang Y., Yang Y. Multihop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries // arXiv:2401.15391. 2024. 10. https://doi.org/10.48550/arXiv.2401.15391
11. Yang X. et al. CRAG – Comprehensive RAG Benchmark // Advances in Neural Information Processing Systems. 2024. Vol. 37. P. 10470–10490. 11. https://doi.org/10.52202/079017-0335
12. Pipitone N., Alami G.H. LegalBench-RAG: A Benchmark for Retrieval-Augmented Generation in the Legal Domain // arXiv:2408.10343. 2024. 12. https://doi.org/10.48550/arXiv.2408.10343
13. Es S., James J., Espinosa Anke L., Schockaert S. RAGAs: Automated Evaluation of Retrieval Augmented Generation // Proc. of the 18th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations. St. Julians, Malta, March 2024. P. 150–158. 13. https://doi.org/10.18653/v1/2024.eacl-demo.16
14. Saad-Falcon J., Khattab O., Potts C., Zaharia M. ARES: An Automated Evaluation Framework for Retrieval-Augmented Generation Systems // Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Vol. 1: Long Papers). Mexico City, Mexico, June 2024. P. 338–354. https://doi.org/10.18653/v1/2024.naacl-long.20
15. Friel R., Belyi M., Sanyal A. RAGBench: Explainable Benchmark for Retrieval-Augmented Generation Systems // arXiv:2407.11005. 2024. 15. https://doi.org/10.48550/arXiv.2407.11005
16. Wang Z., et al. RAGRouter-Bench: A Dataset and Benchmark for Adaptive RAG Routing // arXiv:2602.00296. 2026. https://doi.org/10.48550/arXiv.2602.00296
17. Bao F.S., et al. FaithBench: A Diverse Hallucination Benchmark for Summarization by Modern LLMs // Proc. of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Vol. 2: Short Papers). Albuquerque, New Mexico, April 2025. P. 448–461. https://doi.org/10.18653/v1/2025.naacl-short.38
18. Tamber M.S., et al. Benchmarking LLM Faithfulness in RAG with Evolving Leaderboards // Proc. of the 2025 Conference on Empirical Methods in Natural Language Processing: Industry Track. Suzhou (China), November 2025. P. 799–811. 18. https://doi.org/10.18653/v1/2025.emnlp-industry.54
19. Efimov P., Chertok A., Boytsov L., Braslavski P. SberQuAD–Russian Reading Comprehension Dataset: Description and Analysis // 11th Conference and Labs of the Evaluation Forum, CLEF 2020. Springer Science and Business Media Deutschland GmbH, 2020. P. 3–15. https://doi.org/10.1007/978-3-030-58219-7_1
20. Vatolin A., Gerasimenko N., Ianina A., Vorontsov K. RuSciBench: Open Benchmark for Russian and English Scientific Document Representations // Doklady Mathematics. 2024. Vol. 110. P. S251–S260. 20. https://doi.org/10.1134/S1064562424602191
21. Chen J., et al. AIR-Bench: Automated Heterogeneous Information Retrieval Benchmark // Proc. of the 63rd Annual Meeting of the Association for Computational Linguistics (Vol. 1: Long Papers). Vienna, Austria, July 2025. P. 19991–20022. https://doi.org/10.18653/v1/2025.acl-long.982.
22. Chen J., et al. M3-Embedding: Multilinguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillation // Findings of the Association for Computational Linguistics: ACL 2024. Bangkok, Thailand, August 2024. P. 2318–2335. https://doi.org/10.18653/v1/2024.findings-acl.137
23. Kovalev G., Tikhomirov M., Kozhevnikov E., Kornilov M., Loukachevitch N. Building Russian Benchmark for Evaluation of Information Retrieval Models // arXiv:2504.12879. 2025. https://doi.org/10.48550/arXiv.2504.12879
24. Kwon W., et al. Efficient Memory Management for Large Language Model Serving with PagedAttention // Proc. of the 29th Symposium on Operating Systems Principles, SOSP '23. New York, NY, USA: Association for Computing Machinery, 2023. P. 611–626. https://doi.org/10.1145/3600006.3613165
25. Tkachenko M., Malyuk M., Holmanyuk A., Liubimov N. Label Studio: Data Labeling Software. 2020–2025. URL: https://github.com/HumanSignal/label-studio (Accessed 20.08.2026).
26. Lin C.-Y. ROUGE: A Package for Automatic Evaluation of Summaries // Text Summarization Branches Out. Barcelona, Spain, July 2004. P. 74–81. URL: https://aclanthology.org/W04-1013/ (Accessed: 20.08.2026).
This work is licensed under a Creative Commons Attribution 4.0 International License.
Presenting an article for publication in the Russian Digital Libraries Journal (RDLJ), the authors automatically give consent to grant a limited license to use the materials of the Kazan (Volga) Federal University (KFU) (of course, only if the article is accepted for publication). This means that KFU has the right to publish an article in the next issue of the journal (on the website or in printed form), as well as to reprint this article in the archives of RDLJ CDs or to include in a particular information system or database, produced by KFU.
All copyrighted materials are placed in RDLJ with the consent of the authors. In the event that any of the authors have objected to its publication of materials on this site, the material can be removed, subject to notification to the Editor in writing.
Documents published in RDLJ are protected by copyright and all rights are reserved by the authors. Authors independently monitor compliance with their rights to reproduce or translate their papers published in the journal. If the material is published in RDLJ, reprinted with permission by another publisher or translated into another language, a reference to the original publication.
By submitting an article for publication in RDLJ, authors should take into account that the publication on the Internet, on the one hand, provide unique opportunities for access to their content, but on the other hand, are a new form of information exchange in the global information society where authors and publishers is not always provided with protection against unauthorized copying or other use of materials protected by copyright.
RDLJ is copyrighted. When using materials from the log must indicate the URL: index.phtml page = elbib / rus / journal?. Any change, addition or editing of the author's text are not allowed. Copying individual fragments of articles from the journal is allowed for distribute, remix, adapt, and build upon article, even commercially, as long as they credit that article for the original creation.
Request for the right to reproduce or use any of the materials published in RDLJ should be addressed to the Editor-in-Chief A.M. Elizarov at the following address: amelizarov@gmail.com.
The publishers of RDLJ is not responsible for the view, set out in the published opinion articles.
We suggest the authors of articles downloaded from this page, sign it and send it to the journal publisher's address by e-mail scan copyright agreements on the transfer of non-exclusive rights to use the work.