Contemporaneous Bridge Studies on Foundation-Model Research Integrity, Agentic Science, and Collective Measurement: Eight Working Notes (2025-2026)

This restricted bundle contains eight bridge studies engaging empirical and theoretical developments in foundation-model research published between 2024 and 2026, relating each to a methodological programme on cross-corpus terminology engineering. The studies address: the 2024 Journal of Economic Surveys state-of-the-field paper on metaresearch and its structural parallels to cross-corpus methodology; a methodological pluralism analysis comparing human-led cross-corpus approaches to the agentic-science paradigm; the documented incidence of fabricated citations in recent machine-learning conference proceedings and the argument that cross-corpus audit provides a structural response to the semantic hallucination layer that automated citation verification does not address; a framework on the gradual shift of cognitive load to AI systems and the operationalisation of its three criteria; a critique of agentic-scientist structural limitations and an alternative methodological architecture; EU AI Act (Regulation 2024/1689) regulatory propagation in multi-agent systems and predicted compliance cascade phenomena; distributed sycophancy cascade in multi-agent multi-human teams; and cross-team output convergence and homogenization in foundation-model-assisted knowledge work. All bridge studies are independent research hypotheses produced without institutional affiliation or peer review, shared under CC BY-NC-ND 4.0. Part of the AUGMANITAI Research Programme working corpus (concept-DOI 10.5281/zenodo.20161494). Living document; prepared with AI assistance. AI-transparency (EU AI Act, Art. 50): prepared with substantial large-language-model assistance; all content reviewed and validated by the human author, who retains sole responsibility.