Eduard Gorbunov

94 papers A* 26A 7Journal 59Unranked 2

Year	Rank	Type	Title / Venue / Authors
2026	J	`jnl`	Byzantine-Robust Optimization under (L CoRR Arman Bolatov, Samuel Horváth, Martin Takác, Eduard Gorbunov
2026	J	`jnl`	Byzantine-Robust and Differentially Private Federated Optimization under Weaker Assumptions. CoRR Rustem Islamov, Grigory Malinovsky, Alexander Gaponov, Aurélien Lucchi, Peter Richtárik, Eduard Gorbunov
2026	J	`jnl`	On the Role of Batch Size in Stochastic Conditional Gradient Methods. CoRR Rustem Islamov, Roman Machacek, Aurélien Lucchi, Antonio Silveti-Falls, Eduard Gorbunov, Volkan Cevher
2025	A*	`conf`	Clipping Improves Adam-Norm and AdaGrad-Norm when the Noise Is Heavy-Tailed. ICML Savelii Chezhegov, Yaroslav Klyukin, Andrei Semenov, Aleksandr Beznosikov, Alexander V. Gasnikov, Samuel Horváth, Martin Takác, Eduard Gorbunov
2025	J	`jnl`	Convergence of Clipped-SGD for Convex (L CoRR Saveliy Chezhegov, Aleksandr Beznosikov, Samuel Horváth, Eduard Gorbunov
2025	J	`jnl`	Differentially Private Clipped-SGD: High-Probability Convergence with Arbitrary Clipping Level. CoRR Saleh Vatan Khah, Savelii Chezhegov, Shahrokh Farahmand, Samuel Horváth, Eduard Gorbunov
2025	J	`jnl`	Distributed learning with compressed gradient differences. Optim. Methods Softw. Konstantin Mishchenko, Eduard Gorbunov, Martin Takác, Peter Richtárik
2025	J	`jnl`	Double Momentum and Error Feedback for Clipping with Fast Rates and Differential Privacy. CoRR Rustem Islamov, Samuel Horváth, Aurélien Lucchi, Peter Richtárik, Eduard Gorbunov
2025	J	`jnl`	EF21 with Bells & Whistles: Six Algorithmic Extensions of Modern Error Feedback. J. Mach. Learn. Res. Ilyas Fatkhullin, Igor Sokolov, Eduard Gorbunov, Zhize Li, Peter Richtárik
2025	J	`jnl`	Intermediate Gradient Methods with Relative Inexactness. J. Optim. Theory Appl. Nikita Kornilov, Mohammad S. Alkousa, Eduard Gorbunov, Fedor S. Stonyakin, Pavel E. Dvurechensky, Alexander V. Gasnikov
2025	A*	`conf`	Methods for Convex (L0, L1)-Smooth Optimization: Clipping, Acceleration, and Adaptivity. ICLR Eduard Gorbunov, Nazarii Tupitsa, Sayantan Choudhury, Alen Aliev, Peter Richtárik, Samuel Horváth, Martin Takác
2025	A*	`conf`	Methods with Local Steps and Random Reshuffling for Generally Smooth Non-Convex Federated Optimization. ICLR Yury Demidovich, Petr Ostroukhov, Grigory Malinovsky, Samuel Horváth, Martin Takác, Peter Richtárik, Eduard Gorbunov
2025	J	`jnl`	On the Interaction of Noise, Compression Role, and Adaptivity under (L CoRR Enea Monzio Compagnoni, Rustem Islamov, Antonio Orvieto, Eduard Gorbunov
2025	J	`jnl`	Partially Personalized Federated Learning: Breaking the Curse of Data Heterogeneity. Trans. Mach. Learn. Res. Konstantin Mishchenko, Rustem Islamov, Eduard Gorbunov, Samuel Horváth
2024	A	`conf`	Breaking the Heavy-Tailed Noise Barrier in Stochastic Optimization Problems. AISTATS Nikita Puchkin, Eduard Gorbunov, Nikolay Kutuzov, Alexander V. Gasnikov
2024	A*	`conf`	Byzantine Robustness and Partial Participation Can Be Achieved at Once: Just Clip Gradient Differences. NeurIPS Grigory Malinovsky, Peter Richtárik, Samuel Horváth, Eduard Gorbunov
2024	A	`conf`	Communication Compression for Byzantine Robust Learning: New Efficient Algorithms and Improved Rates. AISTATS Ahmad Rammal, Kaja Gruntkowska, Nikita Fedin, Eduard Gorbunov, Peter Richtárik
2024	A*	`conf`	Don't Compress Gradients in Random Reshuffling: Compress Gradient Differences. NeurIPS Abdurakhmon Sadiev, Grigory Malinovsky, Eduard Gorbunov, Igor Sokolov, Ahmed Khaled, Konstantin Burlachenko, Peter Richtárik
2024	J	`jnl`	Error Feedback under (L CoRR Sarit Khirirat, Abdurakhmon Sadiev, Artem Riabinin, Eduard Gorbunov, Peter Richtárik
2024	A*	`conf`	Exploring Jacobian Inexactness in Second-Order Methods for Variational Inequalities: Lower Bounds, Optimal Algorithms and Quasi-Newton Approximations. NeurIPS Artem Agafonov, Petr Ostroukhov, Roman Mozhaev, Konstantin Yakovlev, Eduard Gorbunov, Martin Takác, Alexander V. Gasnikov, Dmitry Kamzolov
2024	J	`jnl`	Federated Learning Can Find Friends That Are Beneficial. CoRR Nazarii Tupitsa, Samuel Horváth, Martin Takác, Eduard Gorbunov
2024	J	`jnl`	Gradient Clipping Improves AdaGrad when the Noise Is Heavy-Tailed. CoRR Saveliy Chezhegov, Yaroslav Klyukin, Andrei Semenov, Aleksandr Beznosikov, Alexander V. Gasnikov, Samuel Horváth, Martin Takác, Eduard Gorbunov
2024	J	`jnl`	High-Probability Complexity Bounds for Non-smooth Stochastic Convex Optimization with Heavy-Tailed Noise. J. Optim. Theory Appl. Eduard Gorbunov, Marina Danilova, Innokentiy Shibaev, Pavel E. Dvurechensky, Alexander V. Gasnikov
2024	A*	`conf`	High-Probability Convergence for Composite and Distributed Stochastic Minimization and Variational Inequalities with Heavy-Tailed Noise. ICML Eduard Gorbunov, Abdurakhmon Sadiev, Marina Danilova, Samuel Horváth, Gauthier Gidel, Pavel E. Dvurechensky, Alexander V. Gasnikov, Peter Richtárik
2024	J	`jnl`	Implicitly normalized forecaster with clipping for linear and non-linear heavy-tailed multi-armed bandits. Comput. Manag. Sci. Yuriy Dorn, Nikita Kornilov, Nikolay Kutuzov, Alexander Nazin, Eduard Gorbunov, Alexander V. Gasnikov
2024	J	`jnl`	Initialization using Update Approximation is a Silver Bullet for Extremely Efficient Low-Rank Fine-Tuning. CoRR Kaustubh Ponkshe, Raghav Singhal, Eduard Gorbunov, Alexey Tumanov, Samuel Horváth, Praneeth Vepakomma
2024	—	`conf`	Low-Resource Machine Translation through the Lens of Personalized Federated Learning. EMNLP (Findings) Viktor Moskvoretskii, Nazarii Tupitsa, Chris Biemann, Samuel Horváth, Eduard Gorbunov, Irina Nikishina
2024	J	`jnl`	Low-Resource Machine Translation through the Lens of Personalized Federated Learning. CoRR Viktor Moskvoretskii, Nazarii Tupitsa, Chris Biemann, Samuel Horváth, Eduard Gorbunov, Irina Nikishina
2024	J	`jnl`	Methods for Convex (L CoRR Eduard Gorbunov, Nazarii Tupitsa, Sayantan Choudhury, Alen Aliev, Peter Richtárik, Samuel Horváth, Martin Takác
2024	J	`jnl`	Methods with Local Steps and Random Reshuffling for Generally Smooth Non-Convex Federated Optimization. CoRR Yury Demidovich, Petr Ostroukhov, Grigory Malinovsky, Samuel Horváth, Martin Takác, Peter Richtárik, Eduard Gorbunov
2024	A*	`conf`	Remove that Square Root: A New Efficient Scale-Invariant Version of AdaGrad. NeurIPS Sayantan Choudhury, Nazarii Tupitsa, Nicolas Loizou, Samuel Horváth, Martin Takác, Eduard Gorbunov
2024	J	`jnl`	Remove that Square Root: A New Efficient Scale-Invariant Version of AdaGrad. CoRR Sayantan Choudhury, Nazarii Tupitsa, Nicolas Loizou, Samuel Horváth, Martin Takác, Eduard Gorbunov
2023	A*	`conf`	Accelerated Zeroth-order Method for Non-Smooth Stochastic Convex Optimization Problem with Infinite Variance. NeurIPS Nikita Kornilov, Ohad Shamir, Aleksandr V. Lobanov, Darina Dvinskikh, Alexander V. Gasnikov, Innokentiy Shibaev, Eduard Gorbunov, Samuel Horváth
2023	J	`jnl`	Breaking the Heavy-Tailed Noise Barrier in Stochastic Optimization Problems. CoRR Nikita Puchkin, Eduard Gorbunov, Nikolay Kutuzov, Alexander V. Gasnikov
2023	J	`jnl`	Byzantine Robustness and Partial Participation Can Be Achieved Simultaneously: Just Clip Gradient Differences. CoRR Grigory Malinovsky, Peter Richtárik, Samuel Horváth, Eduard Gorbunov
2023	J	`jnl`	Byzantine-Robust Loopless Stochastic Variance-Reduced Gradient. CoRR Nikita Fedin, Eduard Gorbunov
2023	A*	`conf`	Byzantine-Tolerant Methods for Distributed Variational Inequalities. NeurIPS Nazarii Tupitsa, Abdulla Jasem Almansoori, Yanlin Wu, Martin Takác, Karthik Nandakumar, Samuel Horváth, Eduard Gorbunov
2023	J	`jnl`	Byzantine-Tolerant Methods for Distributed Variational Inequalities. CoRR Nazarii Tupitsa, Abdulla Jasem Almansoori, Yanlin Wu, Martin Takác, Karthik Nandakumar, Samuel Horváth, Eduard Gorbunov
2023	J	`jnl`	Clip21: Error Feedback for Gradient Clipping. CoRR Sarit Khirirat, Eduard Gorbunov, Samuel Horváth, Rustem Islamov, Fakhri Karray, Peter Richtárik
2023	J	`jnl`	Communication Compression for Byzantine Robust Learning: New Efficient Algorithms and Improved Rates. CoRR Ahmad Rammal, Kaja Gruntkowska, Nikita Fedin, Eduard Gorbunov, Peter Richtárik
2023	A*	`conf`	Convergence of Proximal Point and Extragradient-Based Methods Beyond Monotonicity: the Case of Negative Comonotonicity. ICML Eduard Gorbunov, Adrien B. Taylor, Samuel Horváth, Gauthier Gidel
2023	A*	`conf`	High-Probability Bounds for Stochastic Optimization and Variational Inequalities: the Case of Unbounded Variance. ICML Abdurakhmon Sadiev, Marina Danilova, Eduard Gorbunov, Samuel Horváth, Gauthier Gidel, Pavel E. Dvurechensky, Alexander V. Gasnikov, Peter Richtárik
2023	J	`jnl`	High-Probability Bounds for Stochastic Optimization and Variational Inequalities: the Case of Unbounded Variance. CoRR Abdurakhmon Sadiev, Marina Danilova, Eduard Gorbunov, Samuel Horváth, Gauthier Gidel, Pavel E. Dvurechensky, Alexander V. Gasnikov, Peter Richtárik
2023	J	`jnl`	High-Probability Convergence for Composite and Distributed Stochastic Minimization and Variational Inequalities with Heavy-Tailed Noise. CoRR Eduard Gorbunov, Abdurakhmon Sadiev, Marina Danilova, Samuel Horváth, Gauthier Gidel, Pavel E. Dvurechensky, Alexander V. Gasnikov, Peter Richtárik
2023	J	`jnl`	Implicitly normalized forecaster with clipping for linear and non-linear heavy-tailed multi-armed bandits. CoRR Yuriy Dorn, Nikita Kornilov, Nikolay Kutuzov, Alexander Nazin, Eduard Gorbunov, Alexander V. Gasnikov
2023	J	`jnl`	Partially Personalized Federated Learning: Breaking the Curse of Data Heterogeneity. CoRR Konstantin Mishchenko, Rustem Islamov, Eduard Gorbunov, Samuel Horváth
2023	A*	`conf`	Single-Call Stochastic Extragradient Methods for Structured Non-monotone Variational Inequalities: Improved Analysis under Weaker Conditions. NeurIPS Sayantan Choudhury, Eduard Gorbunov, Nicolas Loizou
2023	J	`jnl`	Single-Call Stochastic Extragradient Methods for Structured Non-monotone Variational Inequalities: Improved Analysis under Weaker Conditions. CoRR Sayantan Choudhury, Eduard Gorbunov, Nicolas Loizou
2023	A	`conf`	Stochastic Gradient Descent-Ascent: Unified Theory and New Efficient Methods. AISTATS Aleksandr Beznosikov, Eduard Gorbunov, Hugo Berard, Nicolas Loizou
2023	J	`jnl`	Unified analysis of SGD-type methods. CoRR Eduard Gorbunov
2023	A*	`conf`	Variance Reduction is an Antidote to Byzantines: Better Rates, Weaker Assumptions and Communication Compression as a Cherry on the Top. ICLR Eduard Gorbunov, Samuel Horváth, Peter Richtárik, Gauthier Gidel
2022	A*	`conf`	3PC: Three Point Compressors for Communication-Efficient Distributed Training and a Better Theory for Lazy Aggregation. ICML Peter Richtárik, Igor Sokolov, Elnur Gasanov, Ilyas Fatkhullin, Zhize Li, Eduard Gorbunov
2022	J	`jnl`	3PC: Three Point Compressors for Communication-Efficient Distributed Training and a Better Theory for Lazy Aggregation. CoRR Peter Richtárik, Igor Sokolov, Ilyas Fatkhullin, Elnur Gasanov, Zhize Li, Eduard Gorbunov
2022	J	`jnl`	An Accelerated Method for Derivative-Free Smooth Stochastic Convex Optimization. SIAM J. Optim. Eduard Gorbunov, Pavel E. Dvurechensky, Alexander V. Gasnikov
2022	A*	`conf`	Clipped Stochastic Methods for Variational Inequalities with Heavy-Tailed Noise. NeurIPS Eduard Gorbunov, Marina Danilova, David Dobre, Pavel E. Dvurechenskii, Alexander V. Gasnikov, Gauthier Gidel
2022	J	`jnl`	Clipped Stochastic Methods for Variational Inequalities with Heavy-Tailed Noise. CoRR Eduard Gorbunov, Marina Danilova, David Dobre, Pavel E. Dvurechensky, Alexander V. Gasnikov, Gauthier Gidel
2022	J	`jnl`	Distributed Methods with Absolute Compression and Error Compensation. CoRR Marina Danilova, Eduard Gorbunov
2022	A	`conf`	Extragradient Method: O(1/K) Last-Iterate Convergence for Monotone Variational Inequalities and Connections With Cocoercivity. AISTATS Eduard Gorbunov, Nicolas Loizou, Gauthier Gidel
2022	J	`jnl`	Federated Optimization Algorithms with Random Reshuffling and Gradient Compression. CoRR Abdurakhmon Sadiev, Grigory Malinovsky, Eduard Gorbunov, Igor Sokolov, Ahmed Khaled, Konstantin Burlachenko, Peter Richtárik
2022	A*	`conf`	Last-Iterate Convergence of Optimistic Gradient Method for Monotone Variational Inequalities. NeurIPS Eduard Gorbunov, Adrien B. Taylor, Gauthier Gidel
2022	A*	`conf`	Secure Distributed Training at Scale. ICML Eduard Gorbunov, Alexander Borzunov, Michael Diskin, Max Ryabinin
2022	J	`jnl`	Smooth Monotone Stochastic Variational Inequalities and Saddle Point Problems - Survey. CoRR Aleksandr Beznosikov, Boris T. Polyak, Eduard Gorbunov, Dmitry Kovalev, Alexander V. Gasnikov
2022	A	`conf`	Stochastic Extragradient: General Analysis and Improved Rates. AISTATS Eduard Gorbunov, Hugo Berard, Gauthier Gidel, Nicolas Loizou
2022	J	`jnl`	Stochastic Gradient Descent-Ascent: Unified Theory and New Efficient Methods. CoRR Aleksandr Beznosikov, Eduard Gorbunov, Hugo Berard, Nicolas Loizou
2022	J	`jnl`	Variance Reduction is an Antidote to Byzantines: Better Rates, Weaker Assumptions and Communication Compression as a Cherry on the Top. CoRR Eduard Gorbunov, Samuel Horváth, Peter Richtárik, Gauthier Gidel
2021	J	`jnl`	An accelerated directional derivative method for smooth stochastic convex optimization. Eur. J. Oper. Res. Pavel E. Dvurechensky, Eduard Gorbunov, Alexander V. Gasnikov
2021	J	`jnl`	Distributed and Stochastic Optimization Methods with Gradient Compression and Local Steps. CoRR Eduard Gorbunov
2021	J	`jnl`	EF21 with Bells & Whistles: Practical Algorithmic Extensions of Modern Error Feedback. CoRR Ilyas Fatkhullin, Igor Sokolov, Eduard Gorbunov, Zhize Li, Peter Richtárik
2021	J	`jnl`	Extragradient Method: O(1/K) Last-Iterate Convergence for Monotone Variational Inequalities and Connections With Cocoercivity. CoRR Eduard Gorbunov, Nicolas Loizou, Gauthier Gidel
2021	A	`conf`	Local SGD: Unified Theory and New Efficient Methods. AISTATS Eduard Gorbunov, Filip Hanzely, Peter Richtárik
2021	A*	`conf`	MARINA: Faster Non-Convex Distributed Learning with Compression. ICML Eduard Gorbunov, Konstantin Burlachenko, Zhize Li, Peter Richtárik
2021	J	`jnl`	MARINA: Faster Non-Convex Distributed Learning with Compression. CoRR Eduard Gorbunov, Konstantin Burlachenko, Zhize Li, Peter Richtárik
2021	A*	`conf`	Moshpit SGD: Communication-Efficient Decentralized Training on Heterogeneous Unreliable Devices. NeurIPS Max Ryabinin, Eduard Gorbunov, Vsevolod Plokhotnyuk, Gennady Pekhimenko
2021	J	`jnl`	Moshpit SGD: Communication-Efficient Decentralized Training on Heterogeneous Unreliable Devices. CoRR Max Ryabinin, Eduard Gorbunov, Vsevolod Plokhotnyuk, Gennady Pekhimenko
2021	J	`jnl`	Near-Optimal High Probability Complexity Bounds for Non-Smooth Stochastic Optimization with Heavy-Tailed Noise. CoRR Eduard Gorbunov, Marina Danilova, Innokentiy Shibaev, Pavel E. Dvurechensky, Alexander V. Gasnikov
2021	J	`jnl`	Secure Distributed Training at Scale. CoRR Eduard Gorbunov, Alexander Borzunov, Michael Diskin, Max Ryabinin
2021	J	`jnl`	Stochastic Extragradient: General Analysis and Improved Rates. CoRR Eduard Gorbunov, Hugo Berard, Gauthier Gidel, Nicolas Loizou
2020	A*	`conf`	A Stochastic Derivative Free Optimization Method with Momentum. ICLR Eduard Gorbunov, Adel Bibi, Ozan Sener, El Houcine Bergou, Peter Richtárik
2020	A	`conf`	A Unified Theory of SGD: Variance Reduction, Sampling, Quantization and Coordinate Descent. AISTATS Eduard Gorbunov, Filip Hanzely, Peter Richtárik
2020	A*	`conf`	Linearly Converging Error Compensated SGD. NeurIPS Eduard Gorbunov, Dmitry Kovalev, Dmitry Makarenko, Peter Richtárik
2020	J	`jnl`	Linearly Converging Error Compensated SGD. CoRR Eduard Gorbunov, Dmitry Kovalev, Dmitry Makarenko, Peter Richtárik
2020	J	`jnl`	Local SGD: Unified Theory and New Efficient Methods. CoRR Eduard Gorbunov, Filip Hanzely, Peter Richtárik
2020	J	`jnl`	Recent Theoretical Advances in Non-Convex Optimization. CoRR Marina Danilova, Pavel E. Dvurechensky, Alexander V. Gasnikov, Eduard Gorbunov, Sergey Guminov, Dmitry Kamzolov, Innokentiy Shibaev
2020	A*	`conf`	Stochastic Optimization with Heavy-Tailed Noise via Accelerated Gradient Clipping. NeurIPS Eduard Gorbunov, Marina Danilova, Alexander V. Gasnikov
2020	J	`jnl`	Stochastic Optimization with Heavy-Tailed Noise via Accelerated Gradient Clipping. CoRR Eduard Gorbunov, Marina Danilova, Alexander V. Gasnikov
2020	J	`jnl`	Stochastic Three Points Method for Unconstrained Smooth Minimization. SIAM J. Optim. El Houcine Bergou, Eduard Gorbunov, Peter Richtárik
2019	J	`jnl`	A Unified Theory of SGD: Variance Reduction, Sampling, Quantization and Coordinate Descent. CoRR Eduard Gorbunov, Filip Hanzely, Peter Richtárik
2019	J	`jnl`	Distributed Learning with Compressed Gradient Differences. CoRR Konstantin Mishchenko, Eduard Gorbunov, Martin Takác, Peter Richtárik
2019	A*	`conf`	Near Optimal Methods for Minimizing Convex Functions with Lipschitz $p$-th Derivatives. COLT Alexander V. Gasnikov, Pavel E. Dvurechensky, Eduard Gorbunov, Evgeniya A. Vorontsova, Daniil Selikhanovych, César A. Uribe, Bo Jiang, Haoyue Wang, Shuzhong Zhang, Sébastien Bubeck, Qijia Jiang, Yin Tat Lee, Yuanzhi Li, Aaron Sidford
2019	—	`conf`	On Primal and Dual Approaches for Distributed Stochastic Convex Optimization over Networks. CDC Darina Dvinskikh, Eduard Gorbunov, Alexander V. Gasnikov, Pavel E. Dvurechensky, César A. Uribe
2019	A*	`conf`	Optimal Tensor Methods in Smooth Convex and Uniformly ConvexOptimization. COLT Alexander V. Gasnikov, Pavel E. Dvurechensky, Eduard Gorbunov, Evgeniya A. Vorontsova, Daniil Selikhanovych, César A. Uribe
2018	J	`jnl`	An Accelerated Directional Derivative Method for Smooth Stochastic Convex Optimization. CoRR Pavel E. Dvurechensky, Alexander V. Gasnikov, Eduard Gorbunov
2018	J	`jnl`	An Accelerated Method for Derivative-Free Smooth Stochastic Convex Optimization. CoRR Pavel E. Dvurechensky, Alexander V. Gasnikov, Eduard Gorbunov
2018	A*	`conf`	Stochastic Spectral and Conjugate Descent Methods. NeurIPS Dmitry Kovalev, Peter Richtárik, Eduard Gorbunov, Elnur Gasanov