[1]Decentralized Optimization with Coupled ConstraintsDemyan Yarmoshik, Alexander Rogozin, Nikita Kiselev, Daniil Dorin, Alexander Gasnikov, Dmitry KovalevInternational Conference on Learning Representations (ICLR), 2025 A*arXivCodePosterDOI
[2]ConvNets Landscape Convergence: Hessian-Based Analysis of Matricized NetworksVladislav Meshkov, Nikita Kiselev, Andrey Grabovoy2024 Ivannikov Ispras Open Conference (ISPRAS)CodeDOI
[3]Robust Convergence of Loss Landscapes through Distributional AveragingNikita Kiselev, Vladislav Meshkov, Andrey GrabovoyProceedings of the Institute for System Programming of the RAS, 2026CodeDOI
[4]Decoding visual information from neural signals: image reconstruction based on joint functional magnetic resonance imaging and electroencephalography analysisDaniil Dorin, Nikita Kiselev, Andrey GrabovoyInformatics and Applications, 2026CodeDOI
[5]Sample Size Determination: Posterior Distributions ProximityNikita Kiselev, Andrey GrabovoyComputational Management Science, 2025CodeDOI
[6]Sample Size Determination: Likelihood BootstrappingNikita Kiselev, Andrey GrabovoyComputational Mathematics and Mathematical Physics, 2025CodeDOI
[7]Unraveling the Hessian: A Key to Smooth Convergence in Loss Function LandscapesNikita Kiselev, Andrey GrabovoyDoklady Mathematics, 2024arXivCodeDOI
[8]Forecasting fMRI Images From Video Sequences: Linear Model AnalysisDaniil Dorin, Nikita Kiselev, Andrey Grabovoy, Vadim StrijovHealth Information Science and Systems, 2024CodePosterVideoDOI
[9]Kandinsky 6.0 Video: Foundation Models for Synchronized Video and Audio GenerationKandinsky LabarXiv preprint, 2026arXivCode
[10]Kandinsky 5.0: A Family of Foundation Models for Image and Video GenerationKandinsky LabarXiv preprint, 2025 Top-1 Daily, Weekly & Monthly Paper on Hugging Face (November 2025)arXivCodePoster
[11]Closing the Curvature Gap: Full Transformer HessiansEgor Petrov, Nikita Kiselev, Vladislav Meshkov, Andrey GrabovoyarXiv preprint, 2026arXiv
[12]Curvature-Aligned Probing for Local Loss-Landscape StabilizationNikita Kiselev, Andrey GrabovoyarXiv preprint, 2026arXivCode
[13]A mean-square criterion for loss-landscape convergence based on the leading eigenspace of the HessianNikita Kiselev, Andrey GrabovoyProceedings of the 68th MIPT All-Russian Scientific Conference, 2026DOI
[14]Sufficient sample size and its connection to the loss landscape convergenceNikita Kiselev, Vladislav Meshkov, Andrey GrabovoyAbstracts of the 22nd All-Russian Conference with International Participation “Mathematical Methods of Pattern Recognition”, 2025DOI
[15]Decoding visual stimuli from multimodal brain signalsDaniil Dorin, Nikita Kiselev, Andrey GrabovoyAbstracts of the 22nd All-Russian Conference with International Participation “Mathematical Methods of Pattern Recognition”, 2025DOI
[16]Loss landscape convergence as an indicator of sufficient sample sizeNikita Kiselev, Andrey GrabovoyProceedings of the 67th MIPT All-Russian Scientific Conference, 2025DOI
[17]Determining a sufficient sample size from the posterior distribution of model parametersNikita Kiselev, Andrey GrabovoyProceedings of the 66th MIPT All-Russian Scientific Conference, 2024DOI
[18]Spatial and temporal methods of time series analysisDaniil Dorin, Nikita Kiselev, Andrey GrabovoyProceedings of the 66th MIPT All-Russian Scientific Conference, 2024DOI