AI Paper Atlas

Research map / Learning theory and algorithms / Prediction and decision losses

Finding Most Influential Sets

Lucas Darius Konrad, Nikolas Kuschnig

ICML 2026 · Learning theory and algorithms › Prediction and decision losses · approach: Adaptive testing and experiments · 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
use exact Dinkelbach MIS instead of greedy selectionMIS selection · Optimization exactness (n<=50, k<=3)optimality vs enumeration: always matches the enumerated optimum, while greedy sometimes returns suboptimal setsImproved
warm-start Dinkelbach with adversarial initial eta values (ablation)Dinkelbach MIS · n=10^8, k=10^7iterations: from five to sixNo clear effect
use exact MIS rather than greedy on Mongolia microcredit RCT (decreasing direction)MIS selection · Mongolia microcredit RCTinfluence gapImproved
use exact MIS rather than greedy on benchmark regression datasetsMIS selection · Benchmark regressionsinfluence valueNo clear effect
increase query similarity from low to high in linearized STS regressionLinearized text embedding regression · STS benchmarksensitivity to 5,000-MIS removal: Outside 0.04 -> 0.07/0.02; Romney 0.81 -> 1.43/0.41Hurt
use residualization with cross-fitting on partial linear simulationMIS estimator · Residualized simulation, n in {1000, 5000, 10000}influence value vs oracle and set recoveryMixed
compare Dinkelbach MIS with greedy and influence approximations for runtimeMIS computation · Synthetic grid of (n, k)runtime: often by an order of magnitudeImproved

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