Conditional Independence Testing

Ask about this topic

Overview

Conditional independence tests ask whether two variables are still related once a third set of variables is taken into account, a key question in causal discovery and feature relevance. Common approaches use regression residuals and kernel dependence measures, conditional randomization, or models of conditional distributions; newer methods also calibrate p-values and support sequential or federated testing. Methods are evaluated on synthetic and real-world data, often by checking error control, power, and the quality of causal graphs they recover. A central challenge is maintaining reliable error control without losing power or scalability, especially with small samples, mixed or incomplete data, and imperfect conditional models or embeddings.

Latest papers

CardsList

19 papers1 paper in the last four weeks, down 67% on the four weeks before. 0.0% of all new papers.

  1. Embedding-Bias in Conditional Independence Testing

    Oct 8, 2026Nikolaj Thams, Anton Rask LundborgConditional Independence Testing

    Department of Mathematical Sciences University of Copenhagen Copenhagen, Denmark

  2. Domain-Adapted Diffusion Models for Conditional Independence Testing

    Sep 27, 2026Yanfeng Yang, Junda Zhao, Yijie Gao +8Multi-Source Domain AdaptationDomain Adaptation

    The Graduate University for Advanced Studies, SOKENDAI, Tokyo, Japan · The Institute of Statistical Mathematics, Tokyo, Japan · East China Normal University, Shanghai, China +4

  3. Exploring Sparse Autoencoders in Text-Based Causal Confounding Adjustment

    Sep 1, 2026Mian Zhong, Katherine A. Keith, Anjalie FieldCausal Effect EstimationSparse Autoencoders

    Johns Hopkins University · Williams College / Cohere · Cohere

  4. Embedded Conditional Independence Tests for Large Language Model Generated Text with an Application to German Parliament Speeches

    Sep 1, 2026Marco Simnacher, Georg Keilbar, Benjamin König +2LLM EvaluationPrivacy Leakage in Language Models

    Chair of Statistics Humboldt-Universität zu Berlin Spandauer Str. 1, 10178 Berlin, Germany · Chair Digital Health & Machine Learning Hasso-Plattner-Institut for Digital Engineering Prof.-Dr.-Helmert-Straße 2-3, 14482 Potsdam, Germany

  5. Agent Behavioral Contracts II: Certifying Compositional Reliability Without Assuming Independence

    Aug 13, 2026Varun Pratap Bhardwaj, Garima Singh, Arun Pratap BhardwajAgent Failure AnalysisAI Agent Reliability

    Qualixar / Independent Researcher, India · Independent Researcher, India

  6. Conditional Independence Tests for Constraint-Based Causal Discovery: A Survey

    Aug 11, 2026Pavel Averin, Theodoros Moysiadis, Ioannis KatakisCausal DiscoveryConstraint-Based Causal Discovery

    Department of Computer Science School of Sciences and Engineering University of Nicosia 2417, Nicosia, Cyprus

  7. Causal-TS: A Python Library for Causal Discovery in High-Dimensional and Nonstationary Time Series

    Jul 27, 2026Mohammad FesangharyTime-Series Causal DiscoveryCausal Discovery

    Bloomberg LP, New York, NY, USA

  8. Learning Gaussian Graphical Models from a Glauber Trajectory Without Mixing

    Jun 30, 2026Eric Shen, Tony Wu, Mahbod Majid +1Graph Structure LearningProbabilistic Graphical Models

    MIT

  9. Sequential Kernel-based Conditional Independence Testing via Adaptive Betting

    Jun 17, 2026Zheng He, Danica J. SutherlandKernel MethodsConditional Independence Testing

    Department of Computer Science, University of British Columbia, Vancouver, Canada · Alberta Machine Intelligence Institute, Edmonton, Canada

  10. Fast Nonparametric Conditional Independence Testing via Two-Stage Regression

    Jun 16, 2026Eric V. StroblCausal DiscoveryCausal Structure Learning

    Department of Biomedical Informatics, University of Pittsburgh, Pittsburgh, PA 15213, USA

  11. Fourier Feature Methods for Nonlinear Causal Discovery: FFML Scoring, TRFF Scoring, and FFCI Testing in Mixed Data

    May 7, 2026Joseph D. RamseyFourier Feature EmbeddingsCausal Discovery

    Department of Philosophy Carnegie Mellon University Pittsburgh, PA 15213

  12. PAIR-CI: Calibrated Conditional Independence Testing for Causal Discovery with Incomplete Data

    May 6, 2026Thomas S. Robinson, Ranjit LallPermutation TestingIncomplete Data Imputation

    Department of Methodology London School of Economics and Political Science London, UK · Department of Politics and International Relations University of OxfordMay Oxford, UK

  13. Mixture Proportion Estimation and Weakly-supervised Kernel Test for Conditional Independence

    Apr 8, 2026Yushi Hirose, Akito Narahara, Takafumi KanamoriConditional Independence TestingWeakly Supervised Learning

    Institute of Science Tokyo · Institute of Science Tokyo, RIKEN AIP

  14. The Generalised Kernel Covariance Measure

    Apr 4, 2026Luca Bergen, Dino Sejdinovic, Vanessa DidelezConditional Independence TestingKernel Methods

    Leibniz Institute for Prevention Research and Epidemiology – BIPS, Bremen, Germany · Faculty of Mathematics and Computer Science, University of Bremen, Bremen, Germany · School of Mathematical Sciences and Australian Institute for Machine Learning Adelaide University, Adelaide, Australia

  15. On the Number of Conditional Independence Tests in Constraint-based Causal Discovery

    Mar 23, 2026Marc Franquesa Monés, Jiaqi Zhang, Caroline UhlerConstraint-Based Causal DiscoveryCausal Discovery

    Eric and Wendy Schmidt Center, Broad Institute of MIT and Harvard · Centre de Formació Interdisciplinària Superior, Universitat Politècnica de Catalunya · Laboratory for Information and Decision Systems, Massachusetts Institute of Technology

  16. Federated Causal Discovery Across Heterogeneous Datasets under Latent Confounding

    Mar 5, 2026Maximilian Hahn, Alina Zajak, Dominik Heider +1Causal DiscoveryCausal Discovery with Latent Confounding

    University of Münster, Institute of Medical Informatics, Münster, Germany

  17. Empirically Calibrated Conditional Independence Tests

    Feb 24, 2026Milleno Pan, Antoine de Mathelin, Wesley TanseyConditional Independence TestingStatistical Hypothesis Testing

    Computational Oncology Memorial Sloan Kettering Cancer Center

  18. Valid Feature-Level Inference for Tabular Foundation Models via the Conditional Randomization Test

    Feb 19, 2026Mohamed SalemStatistical Hypothesis TestingFeature Selection

  19. Meta-Dependence in Conditional Independence Testing

    Apr 17, 2025Bijan Mazaheri, Jiaqi Zhang, Caroline UhlerConditional Independence Testing

    Thayer School of Engineering, Dartmouth College, Hanover, New Hampshire, USA · Schmidt Center, Broad Institute of MIT and Harvard, Cambridge, Massachusetts, USA · Laboratory for Information and Decision Systems, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA