AI Paper Atlas

Research map / Label-efficient and robust learning / OOD and anomaly detection

On the Importance of Feature Separability in Predicting Out-Of-Distribution Error

RENCHUNZI XIE, Hongxin Wei, Lei Feng, Yuzhou Cao, Bo An

NeurIPS 2023 · Label-efficient and robust learning › OOD and anomaly detection · approach: Out-of-distribution generalization theory · 1 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
Dispersion Score instead of Frechet on TinyImageNetDispersion Score · TinyImageNet avgR^2: 0.847 -> 0.970Improved
Dispersion Score instead of ProjNorm in evaluation time (WRN-50-2)Dispersion Score · CIFAR-10Ctime (s): 575.888 -> 11.017Improved
Dispersion Score instead of ProjNorm on CIFAR-100Dispersion Score · CIFAR-100C avgR^2: 0.985 -> 0.962Hurt
Dispersion Score instead of ProjNorm on CIFAR-10Dispersion Score · CIFAR-10C avgR^2: 0.947 -> 0.972Improved
Dispersion Score instead of ProjNorm under class imbalanceDispersion Score · imbalanced CIFAR-10CR^2: 0.873 -> 0.953Improved
Dispersion Score instead of ProjNorm under class imbalance (time)Dispersion Score · imbalanced CIFAR-10Ctime (s): 398.972 -> 2.780Improved
add class-size weights to the dispersion score (ablation)Dispersion Score · imbalanced CIFAR-10CR^2: 0.712 -> 0.978Improved
add class-size weights to the dispersion score (ablation)Dispersion Score · imbalanced CIFAR-100CR^2: 0.494 -> 0.952Improved
use ground-truth labels instead of pseudo labels (ablation)Dispersion Score · CIFAR-10 avgR^2: 0.972 -> 0.970No clear effect
use ground-truth labels instead of pseudo labels (ablation)Dispersion Score · CIFAR-100 avgR^2: 0.962 -> 0.950No clear effect
use ground-truth labels for feature normalization under adversarial attack (ablation)Dispersion Score · CIFAR-10 PGD, MSEMSE: 1.359 -> 0.483Improved
use K-means clusters instead of classifier pseudo labels (ablation)Dispersion Score · OOD error predictionR^2Hurt

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