Selective Inference

Follow this topic Ask about this topic

Selective inference asks how to make reliable claims when the same data have influenced both what is studied and what is concluded. Methods include conditioning tests and confidence intervals on the selection event, sample splitting or data thinning, and selective calibration; these are applied to feature selection, clustering, scientific discovery, and adaptive benchmark comparisons. Related selective prediction methods decide when a model should answer or abstain, while resource-aware methods choose which examples or regions need expensive computation. The central tension is to account for complex selection procedures without making inference computationally impractical or so conservative that useful discoveries and predictions are lost.

Latest papers

23 papers5 papers in the last four weeks, up 25% on the four weeks before. 0.0% of all new papers.

BRIDGE: An Acceptance Gate to Block False Improvements in Model Updates, with a Retrospective Clinical Evaluation
Oct 21, 2026Dwi SunaryonoCross-Fitting Model Selection
Department of Informatics, Faculty of Intelligent Electrical and Informatics Technology, Institut Teknologi Sepuluh Nopember, Surabaya 60111, Indonesia
When Does Evidence-Graph Structure Predict Answer Correctness? Label-Free Abstention for Structured Retrieval
Oct 16, 2026Anjan GoswamiSelective Inference Graph-Based Retrieval
SmartInfer, Inc.
Finite-sample certification of coverage regret for budgeted document selection under adaptive stratified auditing
Oct 13, 2026Nora SerdyukovaRegret Minimization Selective Inference
Department of Statistics, University of Concepcion, Chile
Selective Inference for Clustering Under Unknown Covariance
Oct 12, 2026Yiqun T. Chen, Lucy L. GaoSelective Inference Statistical Hypothesis Testing
Departments of Biostatistics and Computer Science, Johns Hopkins University · Department of Statistics and Actuarial Science, University of Waterloo
PSI-SINDy: Post-Selection Inference for Sparse Identification of Nonlinear Dynamics
Oct 8, 2026Ashraful Islam, Shuichi Nishino, Tomohiro Shiraishi +1Sparse Identification of Nonlinear Dynamics Selective Inference
RIKEN · Nagoya University
Selective Inference for Deep Clustering in Latent Spaces
Sep 23, 2026Eina Mizui, Tomohiro Shiraishi, Shunichi Nishino +1Selective Inference Clustering
Nagoya University · RIKEN
How Sensitive Are LLM Leaderboard Claims to Hidden Model Selection?
Sep 23, 2026Chen Yang, Jun ChenLLM Evaluation Selective Inference
Department of Statistics, Texas A&M University
Statistical Testing for Multiple Instance Learning via Selective Inference with Applications to Computational Pathology
Sep 21, 2026Noriaki Hashimoto, Shuichi Nishino, Teruyuki Katsuoka +7Multiple Instance Learning Selective Inference
RIKEN · Nagoya University · Institute of Science Tokyo +2
Encrypt What Matters: When Selective Homomorphic Inference Is Efficient
Sep 8, 2026Ali Backour, Juan Reyes, Jaime Punyed +1Efficient Inference Selective Inference
Massachusetts Institute of Technology Cambridge, MA 02139
ADMIL: Attention-Distilled Multiple Instance Learning for Selective Foundation Model Inference in Pathology
Aug 22, 2026Duncan Stothers, Ren-Chin Wu, William LotterSelective Inference Whole-Slide Image Analysis
Department of Data Science, Dana-Farber Cancer Institute, Boston, MA · Department of Pathology, Dana-Farber Cancer Institute, Boston, MA · Department of Pathology, Brigham and Women’s Hospital, Boston, MA +1
Automatic Statistical Test for Rationally Expressible Algorithms by Selective Inference, with Applications to Feature Selection
Aug 5, 2026Teruyuki Katsuoka, Tomohiro Shiraishi, Shuichi Nishino +1Selective Inference Selection Bias
Nagoya University · RIKEN
CogEEGAgent: Toward Autonomous Cognitive EEG Analysis with Grounded Execution and Selection-Aware Verification
Jul 27, 2026Dengzhe Hou, Lingyu Jiang, Fangzhou Lin +1Electroencephalography Selective Inference
1Graduate School of Information Sciences, Tohoku University · 2Unprecedented-scale Data Analytics Center, Tohoku University · 3Texas A&M University +1
Measurement Under Selection: Decoy-Calibrated Failure Audits for Language Models
Jun 8, 2026Vyzantinos Repantis, Ameya Gawde, Harshvardhan SinghLLM Evaluation Language Model Error Detection
The Educational Approaches to Virtual Reality Technologies Lab, University of Ioannina, Greece · Meta Platforms
Trust but Verify: Prover-Verifier Deliberation for Selective LLM Prediction
May 24, 2026João Sedoc, Baotong Zhang, Dean FosterLLM Answer Verification Selective Prediction
New York University
Post-ADC Inference: Valid Inference After Active Data Collection
May 12, 2026Shuichi Nishino, Tomohiro Shiraishi, Teruyuki Katsuoka +1Selective Inference
Nagoya University · RIKEN
Towards Reliable LLM Evaluation: Correcting the Winner's Curse in Adaptive Benchmarking
May 7, 2026Yang Xu, Jiefu Zhang, Haixiang Sun +3LLM Evaluation Selective Inference
Purdue University · Johns Hopkins University · Purdue University, USA
CHASE: Competing Hypotheses for Ambiguity-Aware Selective Prediction
May 2, 2026Kartik Jhawar, Yuhao Geng, Atul N. Parikh +1Selective Prediction Selective Inference
Institute for Digital Molecular Analytics and Science, NTU, Singapore 636921 · School of Electrical and Electronic Engineering, NTU, Singapore 639798 · School of Materials Science and Engineering, NTU, Singapore 639798 +1
Statistical Testing Framework for Clustering Pipelines by Selective Inference
Mar 19, 2026Yugo Miyata, Tomohiro Shiraishi, Shuichi Nishino +1Statistical Hypothesis Testing Clustering
Nagoya University · RIKEN
Partial Causal Structure Learning for Valid Selective Conformal Inference under Interventions
Mar 2, 2026Amir Asiaee, Kavey Aryan, James P. LongSelective Inference Causal Structure Learning
Department of Biostatistics, Vanderbilt University Medical Center, TN 37203, USA · Department of Informatics, King’s College London, London, UK · Department of Biostatistics, MD Anderson Cancer Center, Houston, TX, USA
Making Interpretable Discoveries from Unstructured Data: A High-Dimensional Multiple Hypothesis Testing Approach
Nov 3, 2025Jacob CarlsonSelective Inference Interpretable ML
Harvard University∗
Spacing Test for Fused Lasso
Sep 17, 2025Rieko Tasaka, Joe SuzukiSelective Inference
Department of Mathematical Sciences, Osaka University Toyonaka, Osaka 560-8531, Japan
Optimized Conformal Selection: Powerful Selective Inference After Data-Dependent Score Construction
Nov 27, 2024Tian Bai, Ying JinConformal Prediction Selective Inference
Department of Mathematics and Statistics, McGill University · Data Science Initiative and Department of Health Care Policy, Harvard University
The Limits of Post-Selection Generalization
Jun 15, 2018Kobbi Nissim, Adam Smith, Thomas Steinke +2Generalization Bounds Selective Inference
Department of Computer Science, Georgetown University. · Computer Science Department, Boston University. · IBM Research – Almaden. +2

Oct 21, 2026J/K move · Enter open · S save

BRIDGE: An Acceptance Gate to Block False Improvements in Model Updates, with a Retrospective Clinical Evaluation

Dwi Sunaryono

Department of Informatics, Faculty of Intelligent Electrical and Informatics Technology, Institut Teknologi Sepuluh Nopember, Surabaya 60111, Indonesia

Cross-Fitting Model Selection Selective Inference

Open paper AskPDF arXiv

Models are updated repeatedly, and an update that looks better on validation data but not on new data is a false improvement that can misguide clinical decisions. The naive rule accepts every positive validation gain and is optimistic when the same data choose and judge the update. BRIDGE is a model-agnostic acceptance gate that decides each update from predictions and the available labels, with fixed settings. Each fold of the data judges the update chosen on the other folds (cross-fitting) to reduce selection bias; acceptance requires the gain's lower bootstrap bound to exceed a margin. The four settings are (A) 11 new datasets, (B) synthetic tasks, (C) 12 retrospective clinical tasks, and (D) 39 supporting datasets. The false-improvement rate falls from 19.1% (naive) to 1.1% (BRIDGE) in (A), 20.7% to 1.7% (B), 20.0% to 0% (C), and 12.5% to 0.5% (D); every 95% confidence interval of the reduction over datasets or tasks excludes zero (not over the clinical sources). An equal-size random draw from the naive acceptances contains more false improvements in (A)-(E) (post hoc, p<0.001). On the OpenML collection (E) of 14 new datasets, the false-improvement rate falls from 25.6% to 1.0%. BRIDGE removes 97% to 100% of the false improvements that the naive rule accepts, and its mean gain per deployed update is 1.8 to 3.1 times the naive rule's. The 95% upper bound on the false-improvement rate is below 5.5%: formal in (B) for independent tasks (3.56%), and conditional on the real data tested (0.33% to 5.40%).