Konstantin Mishchenko

59 papers A* 14A 4Journal 39Unranked 2

Year	Rank	Type	Title / Venue / Authors
2025	J	`jnl`	Analysis of an Idealized Stochastic Polyak Method and its Application to Black-Box Model Distillation. CoRR Robert M. Gower, Guillaume Garrigos, Nicolas Loizou, Dimitris Oikonomou, Konstantin Mishchenko, Fabian Schaipp
2025	J	`jnl`	Distributed learning with compressed gradient differences. Optim. Methods Softw. Konstantin Mishchenko, Eduard Gorbunov, Martin Takác, Peter Richtárik
2025	—	`conf`	Hardware-Aware Parallel Prompt Decoding for Memory-Efficient Acceleration of LLM Inference. EMNLP (Findings) Hao Mark Chen, Wayne Luk, Yiu Ka Fai Cedric, Rui Li, Konstantin Mishchenko, Stylianos I. Venieris, Hongxiang Fan
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
2025	J	`jnl`	Smoothing DiLoCo with Primal Averaging for Faster Training of LLMs. CoRR Aaron Defazio, Konstantin Mishchenko, Parameswaran Raman, Hao-Jun Michael Shi, Lin Xiao
2024	A*	`conf`	Adaptive Proximal Gradient Method for Convex Optimization. NeurIPS Yura Malitsky, Konstantin Mishchenko
2024	J	`jnl`	Hardware-Aware Parallel Prompt Decoding for Memory-Efficient Acceleration of LLM Inference. CoRR Hao Mark Chen, Wayne Luk, Ka Fai Cedric Yiu, Rui Li, Konstantin Mishchenko, Stylianos I. Venieris, Hongxiang Fan
2024	A*	`conf`	Prodigy: An Expeditiously Adaptive Parameter-Free Learner. ICML Konstantin Mishchenko, Aaron Defazio
2024	J	`jnl`	Super-Universal Regularized Newton Method. SIAM J. Optim. Nikita Doikov, Konstantin Mishchenko, Yurii E. Nesterov
2024	A*	`conf`	The Road Less Scheduled. NeurIPS Aaron Defazio, Xingyu Yang, Ahmed Khaled, Konstantin Mishchenko, Harsh Mehta, Ashok Cutkosky
2024	J	`jnl`	The Road Less Scheduled. CoRR Aaron Defazio, Xingyu Yang, Harsh Mehta, Konstantin Mishchenko, Ahmed Khaled, Ashok Cutkosky
2023	J	`jnl`	Adaptive Proximal Gradient Method for Convex Optimization. CoRR Yura Malitsky, Konstantin Mishchenko
2023	J	`jnl`	Convergence of First-Order Algorithms for Meta-Learning with Moreau Envelopes. CoRR Konstantin Mishchenko, Slavomír Hanzely, Peter Richtárik
2023	A*	`conf`	DoWG Unleashed: An Efficient Universal Parameter-Free Gradient Descent Method. NeurIPS Ahmed Khaled, Konstantin Mishchenko, Chi Jin
2023	J	`jnl`	DoWG Unleashed: An Efficient Universal Parameter-Free Gradient Descent Method. CoRR Ahmed Khaled, Konstantin Mishchenko, Chi Jin
2023	A*	`conf`	Learning-Rate-Free Learning by D-Adaptation. ICML Aaron Defazio, Konstantin Mishchenko
2023	J	`jnl`	Learning-Rate-Free Learning by D-Adaptation. CoRR Aaron Defazio, Konstantin Mishchenko
2023	J	`jnl`	Partially Personalized Federated Learning: Breaking the Curse of Data Heterogeneity. CoRR Konstantin Mishchenko, Rustem Islamov, Eduard Gorbunov, Samuel Horváth
2023	J	`jnl`	Prodigy: An Expeditiously Adaptive Parameter-Free Learner. CoRR Konstantin Mishchenko, Aaron Defazio
2023	J	`jnl`	Regularized Newton Method with Global \({\boldsymbol{\mathcal{O}(1/{k}^2)}}\) Convergence. SIAM J. Optim. Konstantin Mishchenko
2023	—	`conf`	Server-Side Stepsizes and Sampling Without Replacement Provably Help in Federated Optimization. DistributedML@CoNEXT Grigory Malinovsky, Konstantin Mishchenko, Peter Richtárik
2023	J	`jnl`	Stochastic distributed learning with gradient quantization and double-variance reduction. Optim. Methods Softw. Samuel Horváth, Dmitry Kovalev, Konstantin Mishchenko, Peter Richtárik, Sebastian U. Stich
2023	A*	`conf`	Two Losses Are Better Than One: Faster Optimization Using a Cheaper Proxy. ICML Blake E. Woodworth, Konstantin Mishchenko, Francis R. Bach
2023	J	`jnl`	Two Losses Are Better Than One: Faster Optimization Using a Cheaper Proxy. CoRR Blake E. Woodworth, Konstantin Mishchenko, Francis R. Bach
2023	J	`jnl`	When, Why and How Much? Adaptive Learning Rate Scheduling by Refinement. CoRR Aaron Defazio, Ashok Cutkosky, Harsh Mehta, Konstantin Mishchenko
2022	J	`jnl`	Adaptive Learning Rates for Faster Stochastic Gradient Methods. CoRR Samuel Horváth, Konstantin Mishchenko, Peter Richtárik
2022	A*	`conf`	Asynchronous SGD Beats Minibatch SGD Under Arbitrary Delays. NeurIPS Konstantin Mishchenko, Francis R. Bach, Mathieu Even, Blake E. Woodworth
2022	J	`jnl`	Asynchronous SGD Beats Minibatch SGD Under Arbitrary Delays. CoRR Konstantin Mishchenko, Francis R. Bach, Mathieu Even, Blake E. Woodworth
2022	J	`jnl`	Dualize, Split, Randomize: Toward Fast Nonsmooth Optimization Algorithms. J. Optim. Theory Appl. Adil Salim, Laurent Condat, Konstantin Mishchenko, Peter Richtárik
2022	A*	`conf`	IntSGD: Adaptive Floatless Compression of Stochastic Gradients. ICLR Konstantin Mishchenko, Bokun Wang, Dmitry Kovalev, Peter Richtárik
2022	A*	`conf`	ProxSkip: Yes! Local Gradient Steps Provably Lead to Communication Acceleration! Finally! ICML Konstantin Mishchenko, Grigory Malinovsky, Sebastian U. Stich, Peter Richtárik
2022	J	`jnl`	ProxSkip: Yes! Local Gradient Steps Provably Lead to Communication Acceleration! Finally! CoRR Konstantin Mishchenko, Grigory Malinovsky, Sebastian U. Stich, Peter Richtárik
2022	A*	`conf`	Proximal and Federated Random Reshuffling. ICML Konstantin Mishchenko, Ahmed Khaled, Peter Richtárik
2022	J	`jnl`	Server-Side Stepsizes and Sampling Without Replacement Provably Help in Federated Optimization. CoRR Grigory Malinovsky, Konstantin Mishchenko, Peter Richtárik
2022	J	`jnl`	Super-Universal Regularized Newton Method. CoRR Nikita Doikov, Konstantin Mishchenko, Yurii E. Nesterov
2021	J	`jnl`	IntSGD: Floatless Compression of Stochastic Gradients. CoRR Konstantin Mishchenko, Bokun Wang, Dmitry Kovalev, Peter Richtárik
2021	J	`jnl`	Proximal and Federated Random Reshuffling. CoRR Konstantin Mishchenko, Ahmed Khaled, Peter Richtárik
2021	J	`jnl`	Regularized Newton Method with Global O(1/k CoRR Konstantin Mishchenko
2020	A	`conf`	99% of Worker-Master Communication in Distributed Optimization Is Not Needed. UAI Konstantin Mishchenko, Filip Hanzely, Peter Richtárik
2020	J	`jnl`	A Distributed Flexible Delay-Tolerant Proximal Gradient Algorithm. SIAM J. Optim. Konstantin Mishchenko, Franck Iutzeler, Jérôme Malick
2020	A*	`conf`	Adaptive Gradient Descent without Descent. ICML Yura Malitsky, Konstantin Mishchenko
2020	A	`conf`	DAve-QN: A Distributed Averaged Quasi-Newton Method with Local Superlinear Convergence Rate. AISTATS Saeed Soori, Konstantin Mishchenko, Aryan Mokhtari, Maryam Mehri Dehnavi, Mert Gürbüzbalaban
2020	J	`jnl`	Dualize, Split, Randomize: Fast Nonsmooth Optimization Algorithms. CoRR Adil Salim, Laurent Condat, Konstantin Mishchenko, Peter Richtárik
2020	A*	`conf`	Random Reshuffling: Simple Analysis with Vast Improvements. NeurIPS Konstantin Mishchenko, Ahmed Khaled, Peter Richtárik
2020	J	`jnl`	Random Reshuffling: Simple Analysis with Vast Improvements. CoRR Konstantin Mishchenko, Ahmed Khaled, Peter Richtárik
2020	A	`conf`	Revisiting Stochastic Extragradient. AISTATS Konstantin Mishchenko, Dmitry Kovalev, Egor Shulgin, Peter Richtárik, Yura Malitsky
2020	A	`conf`	Tighter Theory for Local SGD on Identical and Heterogeneous Data. AISTATS Ahmed Khaled, Konstantin Mishchenko, Peter Richtárik
2019	J	`jnl`	99% of Parallel Optimization is Inevitably a Waste of Time. CoRR Konstantin Mishchenko, Filip Hanzely, Peter Richtárik
2019	J	`jnl`	A Self-supervised Approach to Hierarchical Forecasting with Applications to Groupwise Synthetic Controls. CoRR Konstantin Mishchenko, Mallory Montgomery, Federico Vaggi
2019	J	`jnl`	Adaptive gradient descent without descent. CoRR Yura Malitsky, Konstantin Mishchenko
2019	J	`jnl`	Better Communication Complexity for Local SGD. CoRR Ahmed Khaled, Konstantin Mishchenko, Peter Richtárik
2019	J	`jnl`	Distributed Learning with Compressed Gradient Differences. CoRR Konstantin Mishchenko, Eduard Gorbunov, Martin Takác, Peter Richtárik
2019	J	`jnl`	First Analysis of Local GD on Heterogeneous Data. CoRR Ahmed Khaled, Konstantin Mishchenko, Peter Richtárik
2019	J	`jnl`	Revisiting Stochastic Extragradient. CoRR Konstantin Mishchenko, Dmitry Kovalev, Egor Shulgin, Peter Richtárik, Yura Malitsky
2019	J	`jnl`	Sinkhorn Algorithm as a Special Case of Stochastic Mirror Descent. CoRR Konstantin Mishchenko
2019	J	`jnl`	Stochastic Newton and Cubic Newton Methods with Simple Local Linear-Quadratic Rates. CoRR Dmitry Kovalev, Konstantin Mishchenko, Peter Richtárik
2018	A*	`conf`	A Delay-tolerant Proximal-Gradient Algorithm for Distributed Learning. ICML Konstantin Mishchenko, Franck Iutzeler, Jérôme Malick, Massih-Reza Amini
2018	A*	`conf`	SEGA: Variance Reduction via Gradient Sketching. NeurIPS Filip Hanzely, Konstantin Mishchenko, Peter Richtárik
2018	J	`jnl`	SEGA: Variance Reduction via Gradient Sketching. CoRR Filip Hanzely, Konstantin Mishchenko, Peter Richtárik