AI Paper Atlas

Research map / Federated learning and privacy / Communication-efficient FL

Riemannian Federated Learning via Averaging Gradient Streams

Zhenwei Huang, Wen Huang, Pratik Jawanpuria, Bamdev Mishra

ICLR 2026 · Federated learning and privacy › Communication-efficient FL · approach: Local drift correction · 6 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
weight uploaded gradient streams by estimated participation probability (ablation)RFedAGS · PEC on non-IID MNISTsolution qualityImproved
estimate participation probabilities by frequency instead of using true ones (ablation)RFedAGS · PEC on MNISTconvergenceNo clear effect
use arbitrary participation instead of random sampling with equal probability (ablation)RFedAGS · PEC on MNISTconvergenceNo clear effect
set straggler participation probability to 0.05 instead of random sampling (ablation)RFedAGS · PEC with three stragglersCPU timeImproved
increase data heterogeneity level (ablation)RFedAGS · PEC on MNISTsolution qualityHurt
increase number of local updates K (fixed step) (ablation)RFedAGS · PEC on MNISTconvergenceMixed
increase local updates K from 2 to 16 (ablation)RFedAGS · MovieLens 1M LRMC r=5test RMSE: 8.095_{-1} -> 7.654_{-1}Improved
use RFedAGS vs RFedAvg, RFedSVRG, RFedProjRFedAGS · PCA synthetic and CIFAR10accuracy and timeImproved
use RFedAGS vs RFedAvg, RFedSVRGRFedAGS · HSP WordNet and FMC PathMNISTdistance / errorImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy