Technical Paper
4-page or 8-page submissions (excluding references) presenting new methods, systems, benchmarks, datasets, or empirical analyses; evaluated on soundness, novelty, and relevance.
Machine Learning/CORE Unranked
AI4MetaScience 2026
NeurIPS 2026 Workshop on AI for Meta‑Science: Scaling and Organizing Science in the Age of AI
The AI for Meta-Science 2026 workshop, held at NeurIPS 2026 in Paris, explores how AI is transforming the scientific process—including peer review, reproducibility, evaluation, and governance—and aims to establish AI for meta-science as a distinct research field. It brings together researchers from machine learning, philosophy, policy, and science studies to address challenges in scaling scientific quality control amid AI-accelerated research production.
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Position Paper
8-page submissions (excluding references) arguing for a viewpoint on how the scientific ecosystem should change; evaluated on clarity of argument, significance, and engagement with alternative views, not empirical novelty.
Research areas in scope
01
Technical Track
AI/LLM-assisted peer review: methods, evaluation, real-world deployments, and failure modesBenchmarks and evaluation of AI scientists and autonomous research agentsVerification, reproducibility, and quality control of (AI-generated) scientific artifactsDetection of AI-generated content and robustness of review systems against manipulation and gamingReviewer assignment, calibration, and incentivesScience-of-science analyses of AI’s impact on research
02
Position Track
What counts as a publication when AI scientists produce papers at scale?Redesigning peer review, publication criteria, and incentive structures for the AI eraCredit, authorship, and career incentives for human and hybrid human-AI researchGovernance, funding allocation, and institutional responses to AI-accelerated scienceEthics and epistemology of AI in the scientific process, including policies for AI use in reviewing and authoring
Policies worth checking twice
- Submissions are double-blind for novel work and single-blind for previously published work.
- Submissions must use the NeurIPS 2026 LaTeX template with a modified footnote.
- Page limits: 4 pages for concise technical contributions, 8 pages for substantial contributions (excluding references); unlimited appendix allowed.
- Reviewers are only obliged to read the main text, not appendices.
- Dual submission is allowed, provided it complies with the other venue’s policy.
- Previously published work is welcome for presentation.
- Authors are encouraged (but not required) to disclose the role of AI systems in producing the submission.
- AI assistance in reviewing is permitted, but reviewers remain fully responsible for their reviews.
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Last verified September 9, 2026