Sim2Sci 2026

Deadlines
Machine Learning/CORE Unranked

Sim2Sci 2026

Sim2Science: ML with Imperfect Scientific Models

1796940000000Paris, FranceOfficial conference site Site reachable

Sim2Sci 2026 is a NeurIPS workshop focused on machine learning methods that interact with imperfect, misspecified scientific simulators across domains such as chemistry, climate science, neuroscience, and fusion. The workshop aims to build a shared vocabulary and toolkit for leveraging simulators with real-world data, encouraging cross-domain collaboration to mitigate simulator limitations and improve scientific inference.

Official CFP Back to deadlines Verified September 9, 2026

Paper fit

Contribution paths

A strong submission should clearly identify its contribution and evaluate it appropriately.

Workshop Papers

5-page papers (excluding references) on methodological advances, domain-specific applications, analyses, or other contributions bridging simulators or emulators with real-world data using machine learning. Original and early research works are encouraged.

Tiny Papers

2-page extended abstracts (excluding references) aimed at under-represented, under-resourced, and budding researchers. Must address the same topics as workshop papers but in a concise format and meet a high quality bar.

Research areas in scope

01

Topics of Interest

Simulation-based inference and related parameter inference methodsUnderstanding and mitigating model misspecification, including simulator diagnostics and discrepancy modelingEmulator and surrogate modeling, as well as hybrid and physics-informed approachesAnalysis of simulator structure, degeneracy, simplifications, and identifiabilitySimulator pipelines, including data handling, preprocessing, and integration with downstream ML modelsActive learning and Bayesian optimization for fitting parameters or model componentsClosed-loop and experiment-in-the-loop scientific workflowsMulti-fidelity and multi-resolution modeling(Agentic) model and equation discoveryDifferentiable frameworks, LLM-assisted scientific reasoning, and workflow automation

Policies worth checking twice

  • Submissions must use the official NeurIPS 2026 LaTeX template with the dblblindworkshop package option and set the workshop title to Sim2Science.
  • Double-blind review is required; submissions must be anonymized (no author names/affiliations, avoid self-identifying references).
  • Workshop papers are limited to 5 pages and Tiny Papers to 2 pages, excluding references; appendices have no page limit but reviewers are not obligated to read them.
  • Submissions must include the standard NeurIPS reproducibility checklist (placed after references and before appendix), with yes/no/NA answers and brief justifications.
  • Reciprocal reviewing is used: authors must nominate a reviewer with relevant expertise who must accept the invitation and submit adequate reviews; failure results in desk rejection.
  • Submissions that are wholly AI/autonomous-system-generated are not eligible; use of AI writing assistance must follow NeurIPS 2026 policy on LLM use.
  • Workshop is non-archival: accepted work can be submitted elsewhere afterward.
  • Work already published or under review elsewhere is allowed.

Official sources

Compiled from the official call for papers. The organizers’ pages remain authoritative.

Official conference siteProgramCall for Papers

Last verified September 9, 2026