mlsanityData to 14 Oct 2026

Finding Most Influential Sets

arXiv 2606.05919 ↗First posted 4 Jun 2026stat.ML cs.LG econ.EMcited by 0

Not on the map. Why?

Abstract (arXiv)

Identifying most influential sets (MIS) - size-k subsets whose removal maximally changes a target estimand - is typically infeasible because it requires searching over binomnk subsets. For estimands with linear-fractional leave-set-out effects, we show that MIS selection reduces to a one-parameter sequence of top-k problems. Dinkelbach's method yields an algorithm with O(n) cost per iteration and finite termination. For fixed residualized inputs, the algorithm returns a globally optimal set for the univariate ratio objective, including the oracle-residualized partial linear model. With estimated nuisance functions, uniform denominator and generated-score stability imply approximation to the first-order oracle orthogonal-score objective; exact set recovery follows under a separation condition. Simulations and applications show that the method recovers exact MIS that were previously computationally inaccessible.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 15 of its 42 references that are in the map. First posted on arXiv 4 Jun 2026.

Builds on

15 of 42 references.

References · 15

  1. 2017Understanding Black-box Predictions via Influence Functionscited by 1,741
  2. 2017SemEval-2017 Task 1: Semantic Textual Similarity - Multilingual and Cross-lingual Focused Evaluationcited by 1,001
  3. 2020MPNet: Masked and Permuted Pre-training for Language Understandingcited by 981
  4. 2019Data Shapley: Equitable Valuation of Data for Machine Learningcited by 629
  5. 2019On the Accuracy of Influence Functions for Measuring Group Effectscited by 147
  6. 2019On Second-Order Group Influence Functions for Black-Box Predictionscited by 79
  7. 2025Poisoning Attacks on LLMs Require a Near-constant Number of Poison Samplescited by 72
  8. 2024Most Influential Subset Selection: Challenges, Promises, and Beyondcited by 32
  9. 2020An Automatic Finite-Sample Robustness Metric: When Can Dropping a Little Data Make a Big Difference?cited by 27
  10. 2023Data Cleaning and Machine Learning: A Systematic Literature Reviewcited by 9
  11. 2024Approximations to worst-case data dropping: unmasking failure modescited by 8
  12. 2022Provably Auditing Ordinary Least Squares in Low Dimensionscited by 6
3 more
  1. 2023Towards Practical Robustness Auditing for Linear Regressioncited by 6
  2. 2024Robustness Auditing for Linear Regression: To Singularity and Beyondcited by 6
  3. 2025Testing Most Influential Setscited by 2

Built on by

Not cited yet.