This dataset contains a corpus of AI-assisted research artifacts produced between 2025-12-27 and 2026-03-04 as part of the Newton Research Archive. The corpus includes research manuscripts, appendices, replication containers, and supporting materials generated during the development of multiple economics and policy research papers. These materials document a period of intensive AI-assisted research production and provide an empirical basis for studying AI-mediated scholarly workflows. The archive is organized to support reproducibility and traceability. It includes: • Published article PDFs • Replication and appendix containers • Supporting datasets and figures • Index files describing corpus structure Guide files included in the archive: • README. md – overview of the dataset • ARTICLEMAP. md – table of contents with article descriptions • FILEINDEX. txt – full inventory of files in the dataset This dataset accompanies the methodological paper: "Cognitive Trace Publishing (CTP): A Method for Auditable AI-Assisted Research" The corpus illustrates the research workflow environment in which the Cognitive Trace Publishing framework was developed. Related methodological paper: Bell, Peter (2026). "Cognitive Trace Publishing (CTP): A Method for Auditable AI-Assisted Research. "Zenodo. https: //doi. org/10. 5281/zenodo. 18868361 This dataset provides the empirical corpus used to illustrate the research workflow environment in which the Cognitive Trace Publishing framework was developed.
Peter Bell (2026) studied this question.