Pan-Art Pedagogy. Theory & Practice Philology. Theory & Practice Manuscript

Archive of Scientific Articles

ISSUE:    Philology. Theory & Practice. 2026. Volume 19. Issue 9
COLLECTION:    Comparative and Contrastive Research

All issues

License Agreement on scientific materials use.

Comparative experimental study of the interpretation of occasional phraseological units by neural network models based on Russian- and English-language media discourses

Diana Faridovna Khakimullina
Kazan Federal University

Diana Nyailevna Demirag
Kazan Federal University
Submitted: September 11, 2026
Abstract. The presented article is devoted to the problem of hallucinations in large language models (LLMs) when processing complex linguistic units – occasional phraseological units. Despite the widespread prevalence of neural networks in the linguistic field, their responses often demonstrate semantic inadequacy while remaining convincing "on the surface." This phenomenon is referred to as AI hallucination. The aim of the research is to develop a typology of AI hallucinations in the interpretation of occasional phraseological units in Russian- and English-language media discourses. The scientific novelty of the study lies in the fact that, for the first time, a classification of hallucinations specific to occasional phraseological units has been developed and empirically substantiated using Russian and English material. The results obtained showed that the Russian language presents significantly greater difficulty for LLMs than English (an average accuracy of 56% versus 68%), and the typology of hallucinations exhibits a pronounced cross-linguistic specificity: literalization dominates in the English material (40%), whereas prototypicality error (35%) and cultural pseudo-competence (25%) prevail in the Russian material. The proposed typology can be used in developing test suites for evaluating LLMs and in the practice of teaching linguistic disciplines.
Key words and phrases:
галлюцинации ИИ
окказиональные фразеологические единицы
большие языковые модели
медиадискурс
фразеологическая трансформация
AI hallucinations
occasional phraseological units
large language models
media discourse
phraseological transformation
Reader Open the whole article in PDF format. Free PDF-files viewer can be downloaded here.
References:
  1. Айсин Т. Р., Шамардина Т. В. Детекция галлюцинаций на основе внутренних состояний больших языковых моделей // Электронные библиотеки. 2025. Т. 28. Вып. 6. https://doi.org/10.26907/1562-5419-2025-28-6-1282-1305
  2. Батурова Н. И. Окказиональный фразеологизм как языковое явление // Известия Тульского государственного университета. Гуманитарные науки. 2013. № 2.
  3. Кружилина Т. В. К вопросу о моделировании процесса понимания текста с помощью систем искусственного интеллекта // Известия Юго-Западного государственного университета. 2023. Т. 27. № 4. https://doi.org/10.21869/2223-1560-2023-27-4-98-116
  4. Петрова С. И. Окказиональные фразеологизмы в немецкой художественной речи (структурно-семантическая и смысловая характеристики): дисс. ... к. филол. н. М., 1984.
  5. Семушина Е. Ю., Валеева Э. Э. Особенности образования окказиональной расширенной фразеологической метафоры (на материале английского и русского языков) // Филологические науки. Вопросы теории и практики. 2025. № 3. https://doi.org/10.30853/phil20250123
  6. Третьякова И. Ю. Окказиональные фразеологизмы и контекст // Ярославский педагогический вестник. 2011. № 2.
  7. Шалак В. И. Избавление от иллюзий ИИ на примере ChatGPT // Technology and Language. 2024. № 5 (2). https://doi.org/10.48417/technolang.2024.02.03
  8. Шевченко А. А. «Галлюцинации» ИИ как новая форма эпистемической ошибки // Respublica Literaria. 2025. Т. 6. № 4. https://doi.org/10.47850/RL.2025.6.4.93-98
  9. Barassi V. Toward a Theory of AI Errors: Making Sense of Hallucinations, Catastrophic Failures, and the Fallacy of Generative AI // Harvard Data Science Review. 2024. Special Issue 5.
  10. Bender E. M., Koller A. Climbing towards NLU: On Meaning, Form, and Understanding in the Age of Data // Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics (ACL), 2020.
  11. Bottazzi G. E., Ferrario R. The Bewitching AI: The Illusion of Communication with Large Language Models // Philosophy & Technology. 2025. Vol. 38. Iss. 2 (61). https://doi.org/10.1007/s13347-025-00893-6
  12. Ji Z., Lee N., Frieske R., Yu T., Su D., Xu Y., Ishii E., Bang Y. J., Madotto A., Fung P. Survey of Hallucination in Natural Language Generation // ACM Computing Surveys. 2023. Vol. 55. No. 12. https://doi.org/10.1145/3571730
  13. Krakauer D. C., Krakauer J. W., Mitchell M. Large Language Models and Emergence: A Complex Systems Perspective // arXiv preprint. 2025. https://doi.org/10.48550/arXiv.2506.11135
  14. Liu H., Xue W., Chen Y., Chen D., Zhao X., Wang K., Hou L., Li R., Peng W. A Survey on Hallucination in Large Vision-Language Models // arXiv preprint. 2024. https://doi.org/10.20944/preprints202510.0540.v1
  15. Rawte V., Chakraborty S., Pathak A., Sarkar A., Tonmoy S. M. T. I., Chadha A., Sheth A., Das A. The Troubling Emergence of Hallucination in Large Language Models − An Extensive Definition, Quantification, and Prescriptive Remediations // Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2023. https://doi.org/10.18653/v1/2023.emnlp-main.155
  16. Rawte V., Sheth A., Das A. A Survey of Hallucination in Large Foundation Models // arXiv preprint. 2023. https://doi.org/10.48550/arXiv.2309.05922
  17. Sun Y., Sheng D., Zhou Z., Yifei W. AI Hallucination: Towards a Comprehensive Classification of Distorted Information in Artificial Intelligence-Generated Content // Humanities and Social Sciences Communications. 2024. Vol. 11. Article 1278. https://doi.org/10.1057/s41599-024-03811-x
  18. Wang A., Singh A., Michael J., Hill F., Levy O., Bowman S. GLUE: A multi-task benchmark and analysis platform for natural language understanding // Proc. of the 2018 EMNLP Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP, 2018. https://doi.org/10.18653/v1/W18-5446
All issues


© 2006-2025 GRAMOTA Publishing