AI Paper Atlas

Research map / Probabilistic and scientific modeling / Treatment effects

Should I Stop or Should I Go: Early Stopping with Heterogeneous Populations

Hammaad Adam, Fan Yin, Mary Hu, Neil Tenenholtz, Lorin Crawford, Lester Mackey, Allison Koenecke

NeurIPS 2023 · Probabilistic and scientific modeling › Treatment effects · approach: Distributional treatment effects · 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
replace homogeneous stopping test with CLASH (heterogeneous weighting)CLASH · Real A/B testparticipants at stopping: 60,000 -> 40,000Improved
replace SUBTLE with CLASHCLASH · Real A/B testparticipants at stopping: 80,000 -> 40,000Improved
CLASH versus Oracle that only sees harmed regionCLASH · Real A/B testparticipants at stopping: 40,000 -> 40,000No clear effect
apply CLASH to OF-adjusted Cox regressionCLASH · TTE simulation, 25% harmed, HR 0.7stopping probability: +20 percentage pointsImproved
use CLASH instead of homogeneous baseline when minority harmedCLASH · Gaussian simulation, 12.5% minoritystopping probabilityImproved
use CLASH instead of SUBTLE at medium effect sizesCLASH · Gaussian simulationstopping probabilityImproved
use CLASH when treatment has no harmful effectCLASH · Gaussian simulationType I error / stopping probabilityNo clear effect
increase number of covariates (ablation)CLASH · Gaussian simulationstopping probabilityNo clear effect
vary hyperparameter delta (ablation)CLASH · Gaussian simulationstopping probabilityNo clear effect
increase minority group size (ablation)CLASH · Gaussian simulationstopping probabilityImproved
very small sample (N=200) (ablation)CLASH · Gaussian simulationstopping probabilityHurt
very large covariate count (500) (ablation)CLASH · Gaussian simulationstopping probabilityHurt

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