AI Paper Atlas

Research map / Learning theory and algorithms / Prediction and decision losses

On the Power of Source Screening for Learning Shared Feature Extractors

Muxing Wang, Connor Mclaughlin, Lili Su

ICML 2026 · Learning theory and algorithms › Prediction and decision losses · approach: Active learning and subset selection · 0 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 source screening before FedRep trainingFull Population training · ACSIncome (linear)accuracy (%): 72.8 -> 74.2Improved
add source screening before FedRep trainingFull Population training · CelebA smileaccuracy (%): 89.5 -> 90.5Improved
increase assumed dimension k from 3 to 7 in screening (ablation)Ours · ACSIncomeaccuracy (%): 73.4 -> 74.2Improved
increase assumed dimension k from 3 to 7 in screening (ablation)Ours · CelebAaccuracy (%): 90.0 -> 90.5Improved
use random selection of sources instead of screeningFull Population training · ACSIncomeaccuracy (%): 72.8 -> 71.2Hurt
use power-of-choice selectionFull Population training · ACSIncomeaccuracy (%): 72.8 -> 73.0No clear effect
add source screening with two-layer NN extractorFull Population training · ACSIncome (d_h=64)mean accuracy (%): 79.91 -> 80.13Improved
add source screening with two-layer NN extractorFull Population training · ACSIncome (d_h=256)mean accuracy (%): 79.49 -> 79.96Improved
reweight clients by inverse diversity (V1) instead of subsampling (ablation)Source subsampling · ACSIncome (d_h=256)mean accuracy (%): 79.96 -> 79.38Hurt
reweight clients by inverse diversity and data volume (V2) instead of subsampling (ablation)Source subsampling · ACSIncome (d_h=256)mean accuracy (%): 79.96 -> 79.58Hurt
reweight clients (V1) versus full population (ablation)Full Population training · ACSIncome (d_h=256)std of accuracy: 0.06 -> 0.03Improved

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