A focused online event for applying AI to metascience—the study and improvement of how science is done.

Aims

• Clarify AI vs. LLMs and where each fits in metascience.
• Enable ethical, transparent, and effective AI use.
• Provide practical methods for coding, data extraction, and analysis.
• Reduce duplication, improve coordination, and build capacity for funding calls.

Who should attend

• Metascience researchers, methodologists, meta‑research students/postdocs, research software engineers, librarians/information specialists, editors, and research integrity/ethics leads.

Scope

• Plenary: AI capabilities, limits, and evaluation for metascience.
• Workflow: From literature to dataset—deduplication, entity resolution, provenance.
• LLMs in practice: Screening and data extraction with audit trails.
• Beyond LLMs: Retrieval, weak supervision, and classical ML for study classification.
• Reproducible pipelines: Containers, workflow orchestration, and transparency by design.