AI Paper Atlas

Research map / Optimization / Distributed and compressed

Preserved central model for faster bidirectional compression in distributed settings

Constantin Philippenko, Aymeric Dieuleveut

NeurIPS 2021 · Optimization › Distributed and compressed · approach: Compressed distributed optimization · 4 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
add downlink compression on top of uplink compression (MCM vs Diana)Diana · w8alog10 excess loss after 450 epochs: -3.0 -> -2.5Hurt
add downlink compression (MCM vs Diana)Diana · a9alog10 excess loss after 450 epochs: -2.7 -> -2.7No clear effect
add downlink compression (MCM vs Diana)Diana · quantumlog10 excess loss after 450 epochs: -3.2 -> -3.2No clear effect
add downlink compression (MCM vs Diana)Diana · phishinglog10 excess loss after 450 epochs: -3.5 -> -3.4No clear effect
MCM instead of Dore for bidirectional compressionDore · a9alog10 excess loss after 450 epochs: -1.8 -> -2.7Improved
MCM instead of Dore for bidirectional compressionDore · w8alog10 excess loss after 450 epochs: -1.75 -> -2.5Improved
MCM instead of Artemis for bidirectional compressionArtemis · CIFAR-10accuracy after 300 epochs: 54.8 -> 63.5Improved
MCM instead of Dore for bidirectional compressionDore · Fashion MNISTaccuracy after 300 epochs: 87.9 -> 90.6Improved
MCM versus Diana in deep learningDiana · Fashion MNISTaccuracy after 300 epochs: 92.4 -> 90.6Hurt
MCM versus Diana in deep learningDiana · Fashion MNISTtrain loss after 300 epochs: 0.141 -> 0.209Hurt
MCM versus Diana in deep learningDiana · MNISTaccuracy after 300 epochs: 98.9 -> 98.8No clear effect
add randomization of the model (Rand-MCM) (ablation)MCM · least-squares and logistic regressionconvergence stabilityImproved

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