The ChronoCulture branch now rests three claim-notes on a single unrefereed preprint (arXiv:2610.20472) — what independently corroborates it before the branch grows further?
As of 2026-10-17 three claim-notes rest on Saha, Khurana, Huang & Huang (2026), "ChronoCulture: Evaluating Cultural Awareness Across Time in Large Language Models" (arXiv:2610.20472, unrefereed at time of writing, Texas A&M University):
- claim-chronoculture-llms-present-culture-as-static-not-tracking-change
- claim-chronoculture-llm-predictions-track-direction-better-than-magnitude-of-cultural-shift
- claim-chronoculture-naming-older-survey-year-helps-only-as-real-shift-grows-larger
That is the single-source concentration cap (sources.md, Cali 2026-07-29: no more than three claim-notes may rest on a single unrefereed primary until at least one independent source corroborates it). The branch is at its ceiling: the next finding from this preprint does not become a claim-note but is routed here, and the three existing notes stay exactly as they are (seedling, honest, each carrying independent_roots: 1).
The specific verification. The capture itself already names the strongest candidate: Daryani, Bogen & Daepp (2026), "Accurate in space, unreliable in time: How LLMs represent national cultural change" (arXiv:2609.01902) — described by ChronoCulture's own related-work section as concurrent work reaching a consistent conclusion by a different method (the Inglehart–Welzel cultural map, rather than per-question response-distribution comparison). It was found but not independently fetched or verified during the 2026-10-15 capture session. A direct read of that preprint — confirming it is genuinely a different author group reaching the same structural conclusion by an independently-chosen method, not a restatement or a shared-ancestor result — would discharge the cap. Short of that, the paper's own regional/topic breakdowns (Section 4.5) and MCQ-vs-distribution methodology argument, both held back from claim-note status this session (see the capture's not_promoted list), are further leads from the same source and would not discharge the cap even if written up.
Why it matters. The cap exists because the failure mode is invisible one note at a time: each ChronoCulture note is correctly sourced, correctly flagged seedling, correctly honest about resting on one root — while the branch as a whole could still quietly become a structure the vault believes because it is large, before any second independent group has confirmed it. Corroboration discharges the cap (record it in the notes' audit_status and independent_roots, and the branch may grow); contradiction is the more interesting finding and a note in its own right.