Michael Crawshaw
9 papers · 2022–2025 · 3 conferences · across top CS/AI conferences
Achievements
Jump to papers ↓
π
Interdisciplinary Bridge
π
Conference Polyglot
(3)
π
Cross-Pollinator
(13)
πΊοΈ
Taxonomy Completionist
(11)
π§
Keyword Pioneer
Conferences
ICLR (3)
ICML (3)
NIPS (3)
Top co-authors
Keywords
federated learning
(3)
non-convex optimization
(2)
unbounded smoothness
(2)
gradient clipping
(2)
distributed optimization
(1)
dual averaging
(1)
communication efficiency
(1)
sparse linear regression
(1)
nuclear norm regularization
(1)
recurrent neural network
(1)
client participation
(1)
control variate
(1)
statistical recovery
(1)
client subsampling
(1)
periodic participation
(1)
adam algorithm
(1)
stochastic gradient descent
(1)
nonconvex optimization
(1)
composite optimization
(1)
Papers
Constant Stepsize Local GD for Logistic Regression: Acceleration by Instability
ICML 2025
Complexity Lower Bounds of Adaptive Gradient Algorithms for Non-convex Stochastic Optimization under Relaxed Smoothness
ICLR 2025
Local Steps Speed Up Local GD for Heterogeneous Distributed Logistic Regression
ICLR 2025
Provable Benefits of Local Steps in Heterogeneous Federated Learning for Neural Networks: A Feature Learning Perspective
ICML 2024
Federated Learning under Periodic Client Participation and Heterogeneous Data: A New Communication-Efficient Algorithm and Analysis
NIPS 2024
EPISODE: Episodic Gradient Clipping with Periodic Resampled Corrections for Federated Learning with Heterogeneous Data
ICLR 2023
Federated Learning with Client Subsampling, Data Heterogeneity, and Unbounded Smoothness: A New Algorithm and Lower Bounds
NIPS 2023
Robustness to Unbounded Smoothness of Generalized SignSGD
NIPS 2022
Fast Composite Optimization and Statistical Recovery in Federated Learning
ICML 2022