Release

Follow this topic Ask about this topic

Momentum

5 papers in the last four weeks, against 2 the four weeks before. 0.1% of all new papers.

Jul 6Week of Sep 21

Latest papers 19

All topics

The Price of Correlated Tests: How Strict Should a Model Release Gate Be?
Oct 1, 2026Marco PollanenRelease Gating
Department of Mathematics & Statistics Trent University Peterborough, ON, Canada
PEG-Tab: Sampling-Time Record Repair and Release Control for Tabular Synthesis
Sep 30, 2026Pengfei Li, QinYi Liu, Mohammad KhalilSynthetic Tabular Data Release
When Valid Tool Calls Change Meaning: Formation-Consistent Dispatch for LLM Agents
Sep 28, 2026Geonwoo Kim, Brent ByungHoon KangAgentic Deployments Call
Korea Advanced Institute of Science and Technology
Requirement-Bound Verified Commissioning: A Frozen Four-Billion-Parameter Local Model as a Candidate Generator under an External Acceptance Layer with Verification and Release Authority
Sep 24, 2026Mehmet IscanRelease Acceptance
PythaLab, Yıldız Technical University Istanbul, Turkey
Low-Cost Assays for Measuring Model Behavior Across Vendors and Releases
Sep 24, 2026Tapan ParikhRelease Open-Source
Cornell Tech
How Do Agent Harnesses Create Value? Planning Information and Release Control in Stateful LLM Agents
Sep 17, 2026Yukun Zhang, Kemu Xu, Yishen ChenAgent Harness Large Language Model Planning
The Chinese University of Hong Kong, Shenzhen, China · University of Edinburgh, Edinburgh, United Kingdom · The Chinese University of Hong Kong, Hong Kong, China
Decoupling Readiness from Release for Tail-Aware Scheduling of Agentic LLM Workflows
Sep 12, 2026Bochao Feng, Jianjiang Li, Haojie Wang +4Agentic Workflow Design Schedulers
University of Science and Technology Beijing Beijing, China · Qiyuan Laboratory Beijing, China · Tsinghua University Beijing, China
Non-Prehensile Throwing: A Reinforcement Learning Perspective
Sep 1, 2026Abdullah Mustafa, Ryo Hanai, Ixchel G. Ramirez-Alpizar +4Robotic Grasping Real-World Manipulation Tasks
National Institute of Advanced Industrial Science and Technology (AIST), Japan · Graduate School of Fundamental Science and Engineering, Waseda University, Tokyo 169-8555, Japan
Located but Not Releasable: Silent Gate Inversion and Bounded Linear Release
Aug 12, 2026Xining XunRelease Suppression
Tsingjiao Information Science (Beijing) Co., Ltd.
BulkPR-Bench: Benchmarking Queue-Level Governance of Interacting Pull Requests
Aug 3, 2026Zetong Xiong, Qiao Zhao, Jun Zhang +20Pull Requests Release
Baidu Beijing, China
Auditable Release Control for Pedagogical Leakage in LLM Tutors
Aug 1, 2026Nizam KadirData Leakage Release
Science, Mathematics and Technology Singapore University of Technology and Design (SUTD) Singapore
ScoreShield: Differentially Private Release of Similarity Scores
Jul 27, 2026Behrooz Razeghi, Parsa RahimiΔ)$-Differential Privacy Membership Inference Attacks
School of Engineering and Applied Sciences, Harvard University · School of Engineering, École Polytechnique Fédérale de Lausanne (EPFL)
GEMCo: A Validated, Ethically Releasable Proxy for Inaccessible Counselling Data
Jul 26, 2026Philipp Steigerwald, Eric Rudolph, Mara Stieler +3Counseling Conversational Datasets
Faculty of Computer Science, Center for Artificial Intelligence (KIZ) · Technische Hochschule Nürnberg Georg Simon Ohm, Nürnberg, Germany · Faculty of Social Sciences, Institute for E-Counselling
Predicting LLM Safety Before Release by Simulating Deployment
Jul 8, 2026Marcus Williams, Hannah Sheahan, Cameron Raymond +8Pre-Deployment Safety Assessments Large Language Model Safety
OpenAI
CLAP: Closed-Loop Training, Evaluation, and Release Control for Domain Agent Post-training
Jul 2, 2026Fangfei Li, Chenyang Zhao, Long Wang +3Release
MatrixOrigin China · MatrixOrigin USA
Revisiting Privacy Amplification by Subsampling in Selective Release DPSGD
Jun 3, 2026Xiaobo Huang, Fang XieStochastic Gradient Descent Privacy
Guangdong Provincial Key Laboratory of IRADS, Beijing · Normal-Hong Kong Baptist University · Zhuhai, China
SWE-Chain: Benchmarking Coding Agents on Chained Release-Level Package Upgrades
May 14, 2026Man Ho Lam, Chaozheng Wang, Hange Liu +5Software Engineering Coding Agents
The Chinese University of Hong Kong · Independent · ELLIS +3
When Should an AI Workflow Release? Always-Valid Inference for Black-Box Generate-Verify Systems
May 13, 2026Young Hyun Cho, Will Wei SunAgentic Workflow Design Agentic Workflows
Department of Statistics, Purdue University. · Mitch Daniels School of Business, Purdue University.
Structure Liberates: How Constrained Sensemaking Produces More Novel Research Output
May 1, 2026James Mooney, Zae Myung Kim, Young-Jun Lee +1Scientific Ideation Release
University of Minnesota

Oct 1, 2026, stat.MLJ/K move · Enter open · S save

The Price of Correlated Tests: How Strict Should a Model Release Gate Be?

Marco Pollanen

Department of Mathematics & Statistics Trent University Peterborough, ON, Canada

Release Gating Validation Two-Sample Testing Model Evaluation Large Language Model Reliability+1

Open paper AskPDF arXiv

Before a machine learning model ships, it often has to pass a suite of automated tests. Requiring every test to pass looks safe, yet it can reject many models that would have served users well, and it does not say how trustworthy a passing model actually is. We treat the release gate as a design problem: choose how many tests a model must pass so that cleared models meet a stated reliability target, while keeping as many good models as possible. A two-class latent-factor model makes both costs explicit and reduces each calculation to a one-dimensional integral. We prove that when both classes share the same latent correlation, a stricter gate always raises reliability, so the gate that keeps the most good models is the most lenient one that still meets the target. Under pass-all gating, any reliability target short of perfection is attainable within the model, but the share of good models kept tends to zero as the suite grows. Correlation between tests sets the price. In one configuration, a 99 percent target needs 8 independent tests, but 74 tests at a latent correlation of 0.3 and 5,182 at 0.5, where the gate keeps fewer than one good model in ten. We also give a validation procedure, built on exact binomial bounds, that certifies a gate from labelled data even when the gate is chosen from a fixed shortlist.

Momentum

Related topics

Latest papers 19