AKAnna Kalyuzhnaya

Anna Kalyuzhnaya

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Anna Kalyuzhnaya has a total of 113 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Evolving Multi-Agent Systems with Large Language Models: Surrogate Fitness Limitations and Empirical Insights

    by Vadim Akhmerov, Dmitry Gilemkhanov, Jerzy Kamiński, Ivan Dubrovsky, Maria Zaitseva, Dmitrii Kuznechenko, Egor Rozhdestvenskiy, Danil Fedorov, Ilia Revin, Vadim Potemkin, et al.

    We investigate the use of large language models (LLMs) as evolutionary operators for optimizing the architecture of multi-agent systems (MAS) - a representative instance of LLM-driven optimization over attributed-graph artifacts that arise in generative design. Given a task description, an initial MAS configuration is generated and then refined on a per-task basis through two optimization strategies: (1) a population-based evolutionary approach where LLMs serve as mutation, crossover, and selection operators, with an LLM-as-a-Judge ensemble providing the surrogate fitness signal; and (2) a simpler iterative regeneration scheme that uses textual feedback from LLM judges to guide single-solution improvement without maintaining a population. We evaluate both approaches on the GAIA benchmark. Our experiments reveal a critical sensitivity to the judge model's quality: a capable judge (Claude 4 Sonnet) yields +14.3 and +6.67 percentage-point accuracy improvements for evolutionary and iterative approaches, respectively, whereas a weaker judge (Gemini-2.5-Flash) provides no benefit or even degrades performance below the unoptimized baseline. Both approaches incur substantial computational costs, raising questions about practical cost-effectiveness. These results highlight fundamental challenges of surrogate fitness functions in LLM-driven evolutionary optimization and offer practical insights for applying evolutionary methods to MAS architecture design.

  • Open Access

    Too Many Tools, Too Much Confusion? Navigating Agentic Tool Selection at Scale

    by Jerzy Kamiński, Ilya Galyukshev, Sergey Chuprin, Artem Kuznetsov, Anna Kalyuzhnaya

    This paper addresses the critical scalability challenge that large language model agents face when operating over massive tool repositories. As tool catalogs expand to hundreds or thousands of functions, current architectures exhibit substantial performance degradation caused by semantic collisions between similar tools and ineffective handling of complex multi-tool scenarios. To address these bottlenecks, we propose a recall-first Retrieval–Plan–Select (RPS) framework that combines context-aware query decomposition with synthetic tool description augmentation. The proposed approach explicitly separates retrieval, planning, and final selection through step-local candidate generation, while augmented tool descriptions enriched with expanded summaries and synthetic user questions reduce representation collisions in dense embedding spaces. Evaluation across Ultratool, ToolLinkOS, and ToolRet demonstrates that contextual decomposition consistently improves end-to-end recall under large tool catalogs, increasing recall from 0.340 to 0.494 on Ultratool, from 0.208 to 0.323 on ToolLinkOS, and from 0.300 to 0.347 on ToolRet. Description augmentation further improves retrieval quality, increasing Recall@10 from 0.288 to 0.403 and reducing high-similarity semantic collisions by 41.9% at the 0.90 cosine-similarity threshold. The proposed framework highlights that scalable tool use should be approached primarily as a recall-oriented retrieval and planning problem rather than as a flat in-context selection task, providing practical guidance for building large-scale tool-augmented agents over modern API and MCP-based ecosystems.

  • Antitarget profiling of mycotoxins via molecular docking

    by Varvara Tkachenko, Ilia Nikitin, Igor Morgunov, Victor Safronov, Adeliya Leleytner, Anna Kalyuzhnaya, Maxim Fedorov

    Understanding the molecular mechanisms underlying mycotoxin toxicity is crucial for risk assessment and public health protection. In this study, we performed comprehensive molecular docking analysis of 434 structurally diverse mycotoxins against a panel of 44 human protein targets associated with acute toxicity effects. Docking results revealed that 18.2% of all interactions had binding energies \(\le \) −9 kcal/mol. We identified which proteins across the panel were most vulnerable to mycotoxins: acetylcholinesterase, serotonin transporter, and adrenergic receptors (ADRB1 and ADRB2). Analysis identified most promiscuous mycotoxins, including \(\beta \)-asperlicin C and ergotamine, the latter demonstrating the most favorable predicted binding to HTR1B (\(-14.6\) kcal/mol), consistent with experimental data. Class-specific binding pattern analysis revealed that aflatoxins and ochratoxins exhibit broad multi-target activity. These findings provide systematic insights into the molecular basis of mycotoxin toxicity and identify priority targets for future experimental validation and public health monitoring strategies. The dataset containing the affinity data is publicly available at https://github.com/chemagents/mycotoxins-antitargets


  • Open Access

    Towards Explainable Computational Toxicology: Linking Antitargets to Rodent Acute Toxicity

    by Ilia Nikitin, Igor Morgunov, Victor Safronov, Anna Kalyuzhnaya, Maxim Fedorov

    Objectives: One of the major trends in modern computational toxicology is the development of explainable predictive tools. However, the complex nature of the mechanistic representation of biological organisms and the lack of relevant data remain limiting factors. Methods: This work provides a publicly available dataset of 12,654 compounds with mouse intravenous LD50 values, as well as docking scores (Vina-GPU 2.0) against 44 toxicity-associated proteins. NIH and Brenk filters were applied to refine the chemical space. Results: Across the entire protein panel, the human ether-a-go-go–related gene channel (hERG/KCNH2), vasopressin receptor 1A (AVPR1A), the L-type voltage-gated calcium channel Cav1.2 (CACNA1C), the potassium voltage-gated channel subfamily KQT member 1 (KCNQ1) and endothelin receptor A (EDNRA) showed the strongest association with acute toxicity. Statistically significant differences were found in the distribution of LD50 values for compounds that bind antitargets compared with non-binders. Using known bioactive molecules such as anisodamine, butaperazine, soman, and several cannabinoids as examples confirmed the effectiveness of inverse docking for elucidating mechanism of action. Conclusions: The dataset offers a resource to advance transparent, mechanism-aware toxicity modeling. The data is openly available.


  • Open Access

    MADD: Multi-Agent Drug Discovery Orchestra

    by Gleb V. Solovev, Alina B. Zhidkovskaya, Anastasia Orlova, Nina Gubina, Anastasia Vepreva, Rodion Golovinskii, Ilya Tonkii, Ivan Dubrovsky, Ivan Gurev, Dmitry Gilemkhanov, et al.

    Hit identification is a central challenge in early drug discovery, traditionally requiring substantial experimental resources. Recent advances in artificial intelligence, particularly large language models (LLMs), have enabled virtual screening methods that reduce costs and improve efficiency. However, the growing complexity of these tools has limited their accessibility to wet-lab researchers. Multi-agent systems offer a promising solution by combining the interpretability of LLMs with the precision of specialized models and tools. In this work, we present MADD, a multi-agent system that builds and executes customized hit identification pipelines from natural language queries. MADD employs four coordinated agents to handle key subtasks in de novo compound generation and screening. We evaluate MADD across seven drug discovery cases and demonstrate its superior performance compared to existing LLM-based solutions. Using MADD, we pioneer the application of AI-first drug design to five biological targets and release the identified hit molecules. Finally, we introduce a new benchmark of query-molecule pairs and docking scores for over three million compounds to contribute to the agentic future of drug design.


  • Open Access

    Investigation of the possibility of using evolutionary algorithms for conditional generation of attributed graphs

    by I. Yu. Deeva, P. O. Andreeva, E. N. Shikov, A. V. Kalyuzhnaya

    The field of synthetic generation of attributed graphs is actively developing due to advances in generative modeling. However, a key problem of current methods remains the limited diversity of synthesized graphs, due to the dependence on the characteristics of real data used to train generative models. This is a problem because topological properties of graphs and statistical characteristics of attributes critically affect the performance of graph-based machine learning models. In this paper, we test the hypothesis that a combination of evolutionary algorithms and Bayesian networks can provide flexible control over the generation of both graph topology and attributes. The proposed approach includes two key components: evolutionary algorithms to control topological characteristics of the graph (e.g. average vertex degree, clustering coefficient) and Bayesian networks to generate attributes with given statistical parameters such as assortativity or average correlation between attributes. The method allows explicitly setting constraints on graph properties, providing variability independent of the original data. Experiments confirmed that the approach can generate attributed graphs with a wide range of topological characteristics and given statistical parameters of the attributes with sufficiently low generation error. The results demonstrate the promising use of evolutionary and Bayesian methods for conditional graph generation. The main advantage of the approach is the ability to decompose the problem into independent control of topology and attributes, which opens new possibilities for testing machine learning algorithms under controlled conditions. A limitation is the computational complexity of evolutionary optimization, which requires further work to optimize the algorithm. In the future, the method can be extended to generate dynamic graphs and integrate with deep generative models.

  • Top Cited Papers
    Open Access

    LLM Agents for Smart City Management: Enhancing Decision Support Through Multi-Agent AI Systems

    by Anna Kalyuzhnaya, Sergey Mityagin, Elizaveta Lutsenko, Andrey Getmanov, Yaroslav Aksenkin, Kamil Fatkhiev, Kirill Fedorin, Nikolay O. Nikitin, Natalia Chichkova, Vladimir Vorona, et al.

    This study investigates the implementation of LLM agents in smart city management, leveraging both the inherent language processing abilities of LLMs and the distributed problem solving capabilities of multi-agent systems for the improvement of urban decision making processes. A multi-agent system architecture combines LLMs with existing urban information systems to process complex queries and generate contextually relevant responses for urban planning and management. The research is focused on three main hypotheses testing: (1) LLM agents’ capability for effective routing and processing diverse urban queries, (2) the effectiveness of Retrieval-Augmented Generation (RAG) technology in improving response accuracy when working with local knowledge and regulations, and (3) the impact of integrating LLM agents with existing urban information systems. Our experimental results, based on a comprehensive validation dataset of 150 question–answer pairs, demonstrate significant improvements in decision support capabilities. The multi-agent system achieved pipeline selection accuracy of 94–99% across different models, while the integration of RAG technology improved response accuracy by 17% for strategic development queries and 55% for service accessibility questions. The combined use of document databases and service APIs resulted in the highest performance metrics (G-Eval scores of 0.68–0.74) compared to standalone LLM responses (0.30–0.38). Using St. Petersburg’s Digital Urban Platform as a testbed, we demonstrate the practical applicability of this approach to create integrated city management systems with support complex urban decision making processes. This research contributes to the growing field of AI-enhanced urban management by providing empirical evidence of LLM agents’ effectiveness in processing heterogeneous urban data and supporting strategic planning decisions. Our findings suggest that LLM-based multi-agent systems can significantly enhance the efficiency and accuracy of urban decision making while maintaining high relevance in responses.


  • Open Access

    MADD: Multi-Agent Drug Discovery Orchestra

    by Gleb Vitalevich Solovev, Alina Borisovna Zhidkovskaya, Anastasia Orlova, Nina Gubina, Anastasia Vepreva, Rodion Golovinskii, Ilya Tonkii, Ivan Dubrovsky, Ivan Gurev, Dmitry Gilemkhanov, et al.

    Gleb Vitalevich Solovev, Alina Borisovna Zhidkovskaya, Anastasia Orlova, Nina Gubina, Anastasia Vepreva, Rodion Golovinskii, Ilya Tonkii, Ivan Dubrovsky, Ivan Gurev, Dmitry Gilemkhanov, Denis Chistiakov, Timur A. Aliev, Ivan Poddiakov, Galina Zubkova, Ekaterina V. Skorb, Vladimir Vinogradov, Alexander Boukhanovsky, Nikolay Nikitin, Andrei Dmitrenko, Anna Kalyuzhnaya, Andrey Savchenko. Findings of the Association for Computational Linguistics: EMNLP 2025. 2025.


  • Evolutionary Optimization for Inverse Problem in Engineering: The Case Study of Defects Shape Reconstruction

    by Gleb V. Solovev, Alexander Hvatov, Oleg Petrov, Anna Kalyuzhnaya, Alexandra Klimova, Nikolay O. Nikitin