Answers are grounded in this paper and cite their sources.
Executable method enhances source-chain auditability and reproducibility in AI-assisted research.
Version 2 presents an executable provenance method for heterogeneous AI-assisted research programs and applies it to a bounded 41-record revision-audit cohort. It separates public and private projections, distinguishes date and artifact classes, records provider discrepancies without inferring approval, binds generated counts to governed inputs, and preserves exact formal-artifact boundaries. The public package includes path-free JSON and CSV manifests plus a replay validator. Provenance supports chronology, fixity, reproducibility, and navigation; it does not establish theorem truth, empirical validity, priority, authorship consent, or peer review.
No takes yet. Share an insight, caveat, or question.
Micah Blumberg (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: