About Evolutionary Algorithms to Combat the “Hallucinations” of AI Assistants in the “Modelling and Data Analysis” journal

The fourth issue of the "Modeling and Data Analysis" scientific journal (2025 # 4) has been published.

Among the studies devoted to the application of modern methods of modeling and data analysis, the work of specialists from the Faculty of Information Technology of MSUPE is of particular interest, offering a solution to one of the key problems when working with modern intelligent assistants. The proposed tools are particularly relevant for improving the reliability of using large language models. They allow you to automate AI testing for inconsistency and can be used not only in the IT-field, but also to solve applied psychological problems.

In addition to this work, the issue also presents research on:

• Formation of individual learning trajectories based on Markov models.
• Comparison of transformer architectures for time series forecasting.
• Automated planning using linear programming and methods for testing virtual processes.