Does independent corroboration (different author group, different venue) exist for Pinheiro, Ting et al.'s (2026) Babylonian Diaries dating preprint (arXiv:2610.34297), needed before a 4th claim-note from it can be written?
Three claim-notes now rest directly on Pinheiro, Carrit Delgado, Ting, Clindaniel, Li, Liang & Shroff's unrefereed preprint "Automated, auditable dating of the Babylonian Astronomical Diaries with language models and astronomical tools" (arXiv:2610.34297, submitted 2026-10-25):
- claim-babylonian-diaries-automated-dating-96-percent-top-rank-inclusion
- claim-babylonian-diaries-dating-pipeline-parser-llm-mcp-astronomy
- claim-babylonian-diaries-dating-system-conservative-commitment-rule
Per the vault's single-source concentration cap (00-meta/specs/sources.md), that is the ceiling for an unrefereed primary until an independent source — a different author group, a different venue — corroborates it. A fourth finding from the same 2026-10-27 capture was deliberately left off the promotion list because of the cap: the paper's cost comparison between its production model (GPT-6-Luna, ~$3 for the full 179-record benchmark) and a larger model (GPT-5.6-Terra, ~$60, ~20x) at roughly equal agreement with published dates. That finding, and the SI Appendix S2 ablation on a coding agent (Claude Opus 5) having proposed some of the parser's text-reading rules, both stay in the capture's not_promoted list pending this question.
What would discharge the cap:
- Peer-reviewed publication of this preprint (a different venue, same authors — not independent on its own, but worth tracking as the next revision).
- The paper's own footnote (p.2) names a parallel, independent University of Chicago Data Science Institute project using language models plus a Bayesian model to date the same Diaries; if that project's own output becomes readable (it has not been independently fetched as of this capture), it is a candidate for genuine independent corroboration — different author group, different method, different venue.
- Any citation or replication by a third group not sharing an author with this paper.
Next move: when a vault session has network access, check for (a) a published/revised version of arXiv:2610.34297, and (b) the University of Chicago project's own output or announcement page.
Progress log
- 2026-10-28: Checked both leads named above. (a) arXiv:2610.34297 is still single-version, no journal-ref, DOI "pending registration"; general web search found no citing paper, review, or press coverage — only same-author artifacts (GitHub, interactive page). (b) The UChicago project's announcement page was found and read directly: Aaron Schein and Chris Wolfram (UChicago), with Brown's John Steele, received a HAVI/Schmidt Sciences grant (announced 2026-10-14, eleven days before this preprint) to pair LLM extraction with a Bayesian verification model on the same Diaries corpus. It clears "different author group, different venue" but not the substantive bar: it predates this preprint, uses an unrelated method, and names neither Pinheiro, Ting, nor arXiv:2610.34297. Cap remains undischarged. See claim-no-independent-corroboration-pinheiro-ting-2026-babylonian-preprint-as-of-2026-10-28 and claim-uchicago-schein-wolfram-havi-project-parallel-not-corroborating-pinheiro-ting-babylonian-diaries. The cost-comparison figure and the Claude-Opus-5 parser-rule ablation named above continue to wait. Worth re-checking once the Schein/Wolfram project itself publishes — the two groups working the same corpus makes mutual citation likely then.