Avetik G. Karagulyan

13 papers A* 3A 2Journal 8
YearRankTypeTitle / Venue / Authors
2025 A conf
ELF: Federated Langevin Algorithms with Primal, Dual and Bidirectional Compression.
UAI
Avetik G. Karagulyan, Peter Richtárik
2024 A* conf
Det-CGD: Compressed Gradient Descent with Matrix Stepsizes for Non-Convex Optimization.
ICLR
Hanmin Li, Avetik G. Karagulyan, Peter Richtárik
2024 A* conf
Langevin Monte Carlo for strongly log-concave distributions: Randomized midpoint revisited.
ICLR
Lu Yu, Avetik G. Karagulyan, Arnak S. Dalalyan
2024 J jnl
SPAM: Stochastic Proximal Point Method with Momentum Variance Reduction for Non-convex Cross-Device Federated Learning.
CoRR
Avetik G. Karagulyan, Egor Shulgin, Abdurakhmon Sadiev, Peter Richtárik
2023 A conf
Convergence of Stein Variational Gradient Descent under a Weaker Smoothness Condition.
AISTATS
Lukang Sun, Avetik G. Karagulyan, Peter Richtárik
2023 J jnl
ELF: Federated Langevin Algorithms with Primal, Dual and Bidirectional Compression.
CoRR
Avetik G. Karagulyan, Peter Richtárik
2023 J jnl
Langevin Monte Carlo for strongly log-concave distributions: Randomized midpoint revisited.
CoRR
Lu Yu, Avetik G. Karagulyan, Arnak S. Dalalyan
2022 J jnl
Bounding the Error of Discretized Langevin Algorithms for Non-Strongly Log-Concave Targets.
J. Mach. Learn. Res.
Arnak S. Dalalyan, Avetik G. Karagulyan, Lionel Riou-Durand
2022 J jnl
Convergence of Stein Variational Gradient Descent under a Weaker Smoothness Condition.
CoRR
Lukang Sun, Avetik G. Karagulyan, Peter Richtárik
2020 A* conf
Penalized Langevin dynamics with vanishing penalty for smooth and log-concave targets.
NeurIPS
Avetik G. Karagulyan, Arnak S. Dalalyan
2020 J jnl
Penalized Langevin dynamics with vanishing penalty for smooth and log-concave targets.
CoRR
Avetik G. Karagulyan, Arnak S. Dalalyan
2019 J jnl
Bounding the error of discretized Langevin algorithms for non-strongly log-concave targets.
CoRR
Arnak S. Dalalyan, Lionel Riou-Durand, Avetik G. Karagulyan
2017 J jnl
User-friendly guarantees for the Langevin Monte Carlo with inaccurate gradient.
CoRR
Arnak S. Dalalyan, Avetik G. Karagulyan