AI Paper Atlas

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

Diagnosing Failures in Generalization from Task-Relevant Representational Geometry

Chi-Ning Chou, Artem Kirsanov, Yao-Yuan Yang, SueYeon Chung

ICLR 2026 · Label-efficient and robust learning › OOD and anomaly detection · approach: Uncertainty and misclassification detection

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 ID effective manifold dimension and utility instead of ID test accuracy to rank pretrained weightsmodel selection · ImageNet v1/v2 transfer to 9 datasetsOOD prediction accuracy: 37.22% -> 73.02%Improved
prefer v1 weights with higher Deff despite v2 having higher ImageNet accuracyImageNet-pretrained models · OOD linear-probe transferaccuracy: 14 of 20 architecturesImproved
use ID test accuracy as predictor of OOD performance on class-disjoint shiftCIFAR-10 trained CNNs · CIFAR-100 linear probecorrelation with OODNo clear effect
use participation ratio, effective dimension and utility as OOD predictorsCIFAR-10 trained CNNs · CIFAR-100 linear probePearson correlationImproved
use ID test accuracy to predict OOD on corrupted CIFAR-10CCIFAR-10 trained CNNs · CIFAR-10CaccuracyImproved
use geometric manifold compression to predict OOD on corrupted dataCIFAR-10 trained CNNs · CIFAR-10CcorrelationNo clear effect
use numerical rank (Tunnel Effect) as OOD predictorCIFAR-10 trained CNNs · CIFAR-100 and ImageNet linear probecorrelationMixed
use Neural Collapse and logit-based scores as OOD predictorsCIFAR-10 trained CNNs · OOD linear probecorrelationHurt

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