AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

On the Variability of Concept Activation Vectors

Julia Wenkmann, Damien Garreau

ICML 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Post-hoc explanation analysis · 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
increase number of random examples NCAV (logistic regression) · ImageNet, ResNet-50CAV variabilityImproved
increase N for TCAV scoresTCAV score · ImageNet, ResNet-50TCAV score varianceNo clear effect
average over s independent runs with fixed total random samplesmulti-run TCAV · ImageNet, ResNet-50TCAV score varianceImproved
use Difference of Means instead of logistic regressionCAV estimator · UCI Adult tabularvariability at fixed computeImproved
use Difference of Means instead of logistic regressionCAV estimator · compute timewall-clock time: about 46x fasterImproved
use Concept Activation Regions instead of CAVsCAR estimator · compute timewall-clock timeHurt
vary layer, normalization, regularization, initialization, optimizer (ablation)CAV · all modalitiesvariance decay trendNo clear effect
change model architecture or modality (GoogLeNet, ViT, BERT, CLIP, tabular)CAV · four modalitiesvariance scalingNo 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