AI Paper Atlas

Research map / Label-efficient and robust learning / Semi-supervised learning

Semi-Supervised Hypothesis Testing by Betting on Predictions

Yaniv Tenzer, Elad Tolochinsky, Yaniv Romano

ICML 2026 · Label-efficient and robust learning › Semi-supervised learning · approach: Self-training theory · 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
add the imputed e-process (betting on predictions) to the convex combination of baselinesconv(LR Y, LR X, PPI) e-process · label shift simulation, low correlation, N=30powerImproved
add the imputed e-process to the baseline combination with abundant unlabeled dataconv(LR Y, LR X, PPI) e-process · label shift simulation, low correlation, N=135powerImproved
add the imputed e-process when correlation between X and Y is high and N=135baseline combination · label shift simulation, high correlation, N=135powerNo clear effect
add the imputed e-process to the baseline combination under concept shiftconv(LR Y, LR Y|X, PPI) e-process · concept shift simulation, N=135powerImproved
add the imputed e-process to baselines for testing conditional improvement of one LLM over anotherbaseline combination · LLM evaluation, X = dataset identifier (GSM8K vs AQUA-RAT), n=10, N=90powerImproved
add the imputed e-process using LLM-judge annotations as X with few unlabeled samplesbaseline combination · LLM evaluation, label shift, N=10powerImproved
add the imputed e-process using LLM-judge annotations with more unlabeled samplesbaseline combination · LLM evaluation, label shift, N=60powerMixed
use TabPFN in-context predictions in the imputed e-process with M=2 null datasetsconv(LR Y, PPI) e-process · California census private health insurance, 2017 vs 2019, n=50, N=100powerImproved
increase the number of sampled null datasets M (ablation)imputed e-process · simulations (Appendix F.3)powerImproved

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