AI Paper Atlas

Research map / Label-efficient and robust learning / Domain adaptation and generalization

Demystifying Disagreement-on-the-Line in High Dimensions

Donghwan Lee, Behrad Moniri, Xinmeng Huang, Edgar Dobriban, Hamed Hassani

ICML 2023 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Mixture and distillation for domain shift · 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
Use shared-sample disagreement instead of risk as the source-domain signalRandom features regression · CIFAR-10-C, Tiny ImageNet-C, and Camelyon17source-target linear relationshipImproved
Use risk rather than disagreement as the source-domain signalRandom features regression · Tiny ImageNet and Camelyon17source-target linear relationshipHurt
Use independent disagreement rather than shared-sample disagreement (ablation)Random features regression · Ridgeless, overparameterized regimesource-target disagreement linearityNo clear effect
Share the training sample between models (ablation)Random features regression · Ridgeless, overparameterized regimesource-target disagreement linearityImproved
Share model weights rather than training samples (ablation)Random features regression · Overparameterized regimesource-target disagreement linearityHurt
Add ridge regularization (ablation)Independent and shared-sample disagreement · Overparameterized regimeexact source-target linearityHurt
Add ridge regularization (ablation)Independent and shared-sample disagreement · Overparameterized regimeapproximate source-target linearityImproved
Move from the overparameterized to the underparameterized regime (ablation)Independent and shared-sample disagreement · Ridgeless random features regressionsource-target disagreement linearityHurt
Increase feature width toward infinite overparameterization (ablation)Shared-sample disagreement · Ridgeless random features regressiondisagreementImproved
Increase feature width toward infinite overparameterization (ablation)Independent and shared-weight disagreement · Ridgeless random features regressiondisagreementNo clear effect
Allow concept shift between source and target (ablation)Disagreement-on-the-line · Distribution shiftsource-target disagreement linearityNo clear effect

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