AI Paper Atlas

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

CHB: A Diagnostic Toolkit for Hardness-Aware Clustering Evaluation

Walid Durani, Philipp Jahn, Collin Leiber, David B. Hoffmann, Thomas Seidl, Claudia Plant, Christian Böhm

ICML 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Correspondence and causal representation learning

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
move representation from raw to best learned (UMAP/DDC) for density methods on CIFAR-10density clustering · CIFAR-10NMI: 0.033 -> 0.176Improved
move representation from raw to best learned (UMAP/DDC) for density methods on CIFAR-100density clustering · CIFAR-100NMI: 0.016 -> 0.351Improved
add CLIP then UMAP representation pipelineclustering on CIFAR-10 · CIFAR-10Delta NMI: +.713Improved
add CLIP then UMAP representation pipelineclustering on SVHN · SVHNDelta NMI: -.044No clear effect
separability regime improves from A to B to Call clustering methods · CHB regimesmedian agreement: 0.067 -> 0.529 -> 0.704Improved
use deep-pipeline (DDC) instead of Spectral in Regime Bclustering family choice · Regime Bmedian agreement: 0.631 -> 0.695Improved
use Spectral instead of deep-pipeline in Regime Cclustering family choice · Regime Cmedian agreement: 0.727 -> 0.782Improved
use density-based methods in Regime Bclustering family choice · Regime Bmedian agreement: 0.05Hurt
apply mismatch-correcting embeddings without separability rescue (CIFAR/STL/SVHN)clustering on Regime A datasets · CIFAR-10/100, STL-10, SVHNNMIMixed

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