Konstantin Mishchenko

Konstantin Mishchenko

Research Scientist



I’m a research scientist at Samsung AI Center in Cambridge, UK. I like using mathematics to make things work in practice, especially in deep learning applications. Beside doing research, I serve as an Action Editor for TMLR, tweet about interesting papers, and give talks about my research. In 2023, I was lucky to receive the Outsdanding Paper Award together with Aaron Defazio for our work on adaptive methods.

Before joining Samsung, I did a postdoc at Inria Sierra with Alexandre d’Aspremont and Francis Bach. I received my PhD in computer science from KAUST, where I worked under the supervision of Peter Richtárik on optimization theory and its applications in machine learning. In 2020, I interned at Google Brain hosted by Nicolas Le Roux and Courtney Paquette. Prior to that, I obtained my double degree MSc diploma from École Normale Supérieure Paris-Saclay and Paris-Dauphine, and a BSc from Moscow Institute of Physics and Technology.

Note that I’m currently not taking any interns.

  • Optimization
  • Deep learning
  • Federated and distributed learning
  • PhD in Computer Science, 2021


  • MSc in Data Science, 2017

    École normale supérieure Paris-Saclay and Paris-Dauphine

  • BSc in Computer Science and Physics, 2016

    Moscow Institute of Physics and Technology


Research Scientist
Jan 2023 – Present Cambridge, UK
Working on federated learning and embedded AI systems as a member of the Distributed AI team.
Inria Sierra
Dec 2021 – Dec 2022 Paris, France
Conducted research on adaptive, second-order, and distributed optimization.

Recent Posts

DoWG accepted at NeurIPS
Our work on an extension of DoG with weighted gradients got accepted for presentation at NeurIPS this year! If you want to try our method, a pytorch implementation is available on github. I hope to see more papers building upon DoG, DoWG, D-Adaptation, and Prodigy, we have barely scratched the surface on what can be done, and some of these methods are already being used in practice.
DoWG accepted at NeurIPS
ICML Outstanding Paper Award
I’m delighted to share that Aaron Defazio and I received the ICML Outstanding Paper Award for our work on D-Adaptation. The associated github repository of our paper has been quite popular and we are working hard on making extensions that will make adaptive methods even more useful for deep learning. Our first extension, Prodigy, is available on github as well and has been performing even better than D-Adaptation in our experiments. Expect more updates from us pretty soon!
ICML Outstanding Paper Award

Recent Papers

Quickly discover relevant content by filtering publications.
(2023). When, Why and How Much? Adaptive Learning Rate Scheduling by Refinement.

PDF Cite arXiv

(2023). Partially Personalized Federated Learning: Breaking the Curse of Data Heterogeneity.

PDF Cite arXiv

(2023). Two Losses Are Better Than One: Faster Optimization Using a Cheaper Proxy.

PDF Cite arXiv ICML

(2023). Learning-Rate-Free Learning by D-Adaptation.

PDF Cite Code arXiv ICML

(2023). Convergence of First-Order Algorithms for Meta-Learning with Moreau Envelopes.

PDF Cite arXiv

(2022). Super-Universal Regularized Newton Method.

PDF Cite Code Slides arXiv

(2022). Adaptive Learning Rates for Faster Stochastic Gradient Methods.

PDF Cite arXiv

(2022). Asynchronous SGD Beats Minibatch SGD Under Arbitrary Delays.

PDF Cite Code Slides arXiv

(2022). ProxSkip: Yes! Local Gradient Steps Provably Lead to Communication Acceleration! Finally!. ICML.

PDF Cite Code Video arXiv ICML

(2022). Server-Side Stepsizes and Sampling Without Replacement Provably Help in Federated Optimization.

PDF Cite arXiv

(2021). IntSGD: Adaptive Floatless Compression of Stochastic Gradients. ICLR.

PDF Cite Code Poster Slides arXiv ICLR

(2021). Proximal and Federated Random Reshuffling. ICML.

PDF Cite Code Slides Video arXiv ICML

(2020). Random Reshuffling: Simple Analysis with Vast Improvements. NeurIPS.

PDF Cite Code Poster Slides arXiv NeurIPS

(2020). Dualize, Split, Randomize: Toward Fast Nonsmooth Optimization Algorithms. JOTA.

PDF Cite Poster arXiv JOTA

(2019). Adaptive Gradient Descent without Descent. ICML.

PDF Cite Code Poster Slides arXiv ICML Video

(2019). First Analysis of Local GD on Heterogeneous Data.

PDF Cite Slides arXiv NeurIPS

(2019). Tighter Theory for Local SGD on Identical and Heterogeneous Data. AISTATS.

PDF Cite Slides arXiv AISTATS

(2019). MISO is Making a Comeback With Better Proofs and Rates.

PDF Cite arXiv

(2019). DAve-QN: A Distributed Averaged Quasi-Newton Method with Local Superlinear Convergence Rate. AISTATS.


(2019). Revisiting Stochastic Extragradient. AISTATS.

PDF Cite Slides arXiv AISTATS

(2019). Stochastic Distributed Learning with Gradient Quantization and Double Variance Reduction. Optimization Methods and Software.

PDF Cite arXiv

(2019). 99% of Worker-Master Communication in Distributed Optimization Is Not Needed. UAI.

PDF Cite arXiv UAI

(2019). Distributed Learning with Compressed Gradient Differences.

PDF Cite arXiv

(2018). SEGA: Variance Reduction via Gradient Sketching. NeurIPS.

PDF Cite arXiv NIPS

(2018). A Delay-tolerant Proximal-Gradient Algorithm for Distributed Learning. ICML.


(2018). A Distributed Flexible Delay-tolerant Proximal Gradient Algorithm. SIOPT.

PDF Cite arXiv SIAM