Call for Papers

We welcome a wide range of contributions — including new methods, applications, analyses, benchmarks, and position pieces — from essentially any scientific domain, including biology, chemistry, physics, materials science, and climate science, that bring together machine learning and scientific simulators. We especially encourage work that incorporates real-world data, though submissions are not limited to it.

Submission Tracks

Topics of Interest

See the full list of topics on the Home page.

Not sure if your simulator fits? Here are a few examples of the kinds of scientific simulators we have in mind — this list is illustrative, not exhaustive:

• Density-functional theory / exchange–correlation functionals and force fields (chemistry, materials science)
• Molecular dynamics and ab initio quantum chemistry methods
• Core-transport and turbulence simulators for tokamaks and fusion plasmas (e.g., TORAX)
• Biophysical circuit, point-neuron, and mean-field models (neuroscience)
• Climate and weather models with sub-grid parametrizations
• Cosmological N-body and galaxy formation simulations
• Computational fluid dynamics solvers
• Protein structure and biomolecular simulators
• …

Other scientific simulators not listed above are equally welcome, provided the work couples them with machine learning methods.

Contribution Guide

• Format: Use the official NeurIPS 2026 LaTeX template with the dblblindworkshop package option and set the workshop title to Sim2Science, i.e. \usepackage[dblblindworkshop]{neurips_2026} and \workshoptitle{Sim2Science}. This is required for the camera-ready version and recommended already for submission. Submissions not using the required template may be desk-rejected.
• Length: 5 pages (excluding references) for workshop papers, or 2 pages (excluding references) for Tiny Papers
• Appendix: no page limit on supplementary/appendix material, but reviewers are not obligated to read it — your main submission (within the page limit) should be self-contained
• Anonymization: Double-blind review required — please anonymize your submission (no author names/affiliations, avoid self-identifying references); failure to anonymize, or to follow the required template, may lead to desk rejection
• Checklist: submissions must include the standard NeurIPS reproducibility checklist (provided as checklist.tex in the template), placed after the references and before the appendix; it does not count toward the page limit. Answer each item yes/no/NA with a brief justification. Missing or incomplete checklists may lead to desk rejection.
• Submission Portal: OpenReview

Review Process

We use reciprocal reviewing, in line with how NeurIPS and other major ML venues run their review processes:

• Nomination: at submission time, authors nominate one author of the paper — someone with sufficient relevant expertise and publications — to serve as the paper's reciprocal reviewer for the workshop. If no co-author has a strong publication record, authors may instead nominate a co-author with sufficient relevant expertise; this nominee is held to the same reviewing standard as any other reciprocal reviewer (see Participation below).
• Load: each nominated reviewer is assigned 2 other submissions to review.
• Assignment: reviewers are assigned papers based on their expertise; assignments may be adjusted by organizers for coverage or conflicts of interest.
• Participation: if a nominated reviewer fails to submit their reviews by the review deadline, or submits highly insufficient or inappropriate reviews, the corresponding submission is subject to desk rejection.

Reviewing is supplemented by program committee members and organizers as needed. Papers are reviewed double-blind.

Non-archival: The workshop will not have formal proceedings, but accepted submissions will be linked on the workshop webpage.

FAQ

Important Dates

Questions?

For questions about paper submissions, please contact us at sim2science@gmail.com