AI Paper Atlas

Research map / Learning theory and algorithms / Clustering

Error Discovery By Clustering Influence Embeddings

Fulton Wang, Julius Adebayo, Sarah Tan, Diego Garcia-Olano, Narine Kokhlikyan

NeurIPS 2023 · Learning theory and algorithms › Clustering · approach: Clustering approximation algorithms

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 InfEmbed instead of Dominoslice discovery · dcbench natural images, rareprecision@10: 0.4 -> 0.65Improved
use InfEmbed instead of Dominoslice discovery · dcbench natural images, correlationprecision@10: 0.45 -> 0.55Improved
use InfEmbed instead of Dominoslice discovery · dcbench medical images, rareprecision@10: 0.39 -> 0.57Improved
use InfEmbed instead of Dominoslice discovery · dcbench medical time series, noisy labelprecision@10: 0.9 -> 0.81Hurt
use InfEmbed instead of Dominoslice discovery · dcbenchcoherence score: 1219 out of 1235Improved
use InfEmbed instead of Dominoslice discovery · dcbenchlabel homogeneityMixed
use InfEmbed instead of PlaneSpotslice discovery · SpotCheckdiscovery rate: 0.85 -> 0.91Improved
use InfEmbed instead of PlaneSpotslice discovery · SpotCheckfalse discovery rate: 0.07 -> 0.15Hurt
replace K-Means with DBSCAN (ablation)InfEmbed · SpotCheckdiscovery rate: 0.91 -> 0.92No clear effect
replace K-Means with spectral clustering (ablation)InfEmbed · SpotCheckdiscovery rate: 0.91 -> 0.89No clear effect
branching factor B=2 vs B=3 (ablation)InfEmbed-Rule · dcbenchslice qualityNo clear effect
use influence embeddings (gradient-based) instead of last-layer representations onlyslice discovery · analysisslice similarityImproved

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