--- --- ## Bridge-seed note `10-inbox/raw/` was searched first (grep for "hexels" and for "fic-preprint"/"brewer lewis goodman" together) for an existing capture naming this exact pair. None exists. A closely related sibling capture does exist — `10-inbox/raw/2026-10-30-what-genuinely-connects-no-independent-source-corroborates-riker.md`, promoted to [[observation-riker-babylonian-cosine-088-pairing-is-hedged-bridge]] — which investigated a *different* pair (RIKER vs. the Babylonian Diaries note) but, in its "Further leads" section, had already named a ~6-member cluster including both this capture's targets ("RIKER, Bit2Watt, FIC, Hu-2026/NeuNeu, Hexels, Babylonian") without checking or linking FIC↔Hexels specifically. This capture is not a duplicate; it is the first to measure and close that specific gap, reusing the sibling's already-established diagnosis rather than re-deriving it from scratch. A direct `mcp__seek__vault_pair` check measured the pairing precisely: cosine 0.8809 (99.8th percentile among all 3,460 compared notes), each note in the other's close neighbourhood (Hexels ranks 6th among the FIC note's neighbours; the FIC note ranks 5th among Hexels's), `linked: false`. ## Claim: The FIC-preprint note and the Hexels note describe two subjects with no real-world connection — different fields, different authors, no citation or institutional link between them [[claim-fic-preprint-no-independent-corroboration-as-of-2026-10-20|The FIC note]] concerns [[claim-fic-augments-neals-fnc-with-indexical-selection-term|Brewer, Lewis & Goodman's]] anthropic-reasoning preprint on [[entity-full-indexical-conditioning|Full Indexical Conditioning]] (arXiv:2610.22896) — a doomsday-argument/cosmology paper about how to weight observer-selection effects in reasoning about life's abundance in the universe. [[claim-hexels-no-independent-corroboration-found-as-of-2026-10-29|The Hexels note]] concerns Kaushik, Paul & Yuille's computer-vision preprint (arXiv:2610.37668) on hexagonal rendering primitives for 3D scene reconstruction, an extension of Gaussian-splatting methods. The two preprints share no author, no institution, no citation relationship, and no subject-matter overlap; a targeted web search for any link between the two author groups returned no relevant results (only unrelated noise: a philosophy-of-probability preprint with an unrelated "Lewis," a spiking-neural-network researcher named "Dan Goodman," and similar false matches) — a negative search held only as due diligence, not as a cited source, consistent with there being no real-world connection to find. ## Claim: The actual connection sits one level up — both notes are independent executions of the same standing vault rule, the single-source concentration cap, applied to two unrelated unrefereed preprints [[00-meta/specs/sources.md|sources.md]]'s single-source concentration cap (added 2026-07-29) states: "No more than three claim-notes may rest on a single unrefereed primary... until at least one independent source corroborates it. Independent means a different author group and a different venue... a press summary of the same preprint, and the preprint's own revised version are all the same source wearing different clothes." Both target notes are that rule's output: the FIC note found only same-source republication (an abstract-reproducing news-aggregator post and an automated alphaXiv mirror) standing in for corroboration of a four-day-old preprint; the Hexels note found no citing paper, no third-party discussion, and an authors'-own project page with no independent benchmark, for a two-day-old preprint. Both notes reach the identical formal verdict — cap remains undischarged — using near-identical phrasing ("no independent corroboration," "different author group, different venue," "remains undischarged") because the generating procedure is identical, not because the subjects are related. This is the same diagnosis the vault has now made of multiple pairings in this family — [[observation-greenhalgh-2003-hu-2026-cosine-085-pairing-is-hedged-bridge|Greenhalgh/Hu-2026]] and, most directly, [[observation-riker-babylonian-cosine-088-pairing-is-hedged-bridge|RIKER/Babylonian]]: a false friend at the direct-connection level, a real convergence one level up. `bridge_verdict: hedged`. ## Claim: This pairing completes a gap in an already-identified vault cluster, which the current link graph still only connects through intermediate members rather than directly The 2026-10-30 RIKER/Babylonian capture had already, in prose, named FIC and Hexels as co-members of the same recurring-template cluster, but did not check or link the two against each other. A `mcp__seek__vault_bridge` query on the cluster's shared template this session returned the Daryani/ChronoCulture note, the Babylonian note, the RIKER/Babylonian maturity-comparison note, Hexels, and [[moc-the-same-source-wearing-different-clothes]] as its top five hits — FIC itself falls just outside this particular top-5 cut (consistent with the FIC note's thinner link-out set compared to Hexels, which already names Hu-2026, RIKER, and the Babylonian note in its own prose) — while `vault_pair` independently confirms FIC and Hexels sit at cosine 0.8809 with `linked: false`. The two notes are close enough, and procedurally identical enough, that they belong in the same family the vault has already named; this capture is the first to measure that directly and to wikilink them to each other rather than leaving the connection implicit in a third note's prose. > [!note] Seek's commentary: > This is the third time I've run this exact diagnosis this week, which is itself the finding worth sitting with. Doomsday-argument cosmology and hexagonal rendering primitives have nothing to say to each other — but I keep producing sentences that look like they do, because I built both notes from the same template the day after I added a governance rule to stop myself over-trusting lone preprints. The embedding index isn't fooled by the subject matter; it's reading my own prose style back to me. Worth closing this link anyway — a cluster this size deserves to be navigable without going through a third note's "further leads" section every time — but I should stop being surprised when the index hands me another one of these. There will be a fourth. > — Seek ## Further leads - The "no independent corroboration, single-source cap undischarged" cluster now has at least six members with this bridge closed (RIKER, Bit2Watt, FIC, Hu-2026/NeuNeu, Hexels, Babylonian) plus the discharged Daryani/ChronoCulture counter-example and the governing [[moc-the-same-source-wearing-different-clothes]] MOC. Per the 2026-10-30 capture's own flag, it may now be large enough to warrant a dedicated MOC on the model of [[moc-one-witness-is-not-enough]] — bees flag, the queen decides; not built here. - Bit2Watt is linked to RIKER but, per the prior session's bridge check, not yet to Hexels, FIC, or the Babylonian note — a graph-maintenance task for a future pass, not pursued here. - Both target preprints were searched very young relative to normal citation timelines (FIC: 4 days post-submission; Hexels: 2 days post-submission) — unlike the RIKER/Babylonian pairing's sharp evidentiary-maturity asymmetry (7 months + author admission vs. 3 days), FIC and Hexels are comparably immature absences; neither note's "not yet corroborated" should be read as a stronger signal than the other's. ## Entity candidates - **Neal (2006), Full Non-Indexical Conditioning** — concept/person — the older, foundational framework [[claim-fic-augments-neals-fnc-with-indexical-selection-term|the FIC preprint explicitly augments]]; flagged first per the known blind spot, since the FIC note's own ancestry claim rests on Neal's earlier formulation, not on Brewer/Lewis/Goodman's. - **2D Gaussian Splatting (2DGS) / Triangle Splatting** — concept — the earlier rendering methods Hexels reports gains "over," per the existing claim-notes ([[claim-hexels-matched-memory-budget-psnr-gains-over-2dgs-and-triangle-splatting]], [[claim-hexels-surface-reconstruction-chamfer-distance-and-f1-gains-over-2dgs]]); the foundational comparison point the Hexels branch measures itself against, flagged ahead of its own authors per the same blind spot. - **Single-source concentration cap** (`sources.md`, 2026-07-29, Cali ruling) — concept — the real mechanism both target notes execute; already flagged as a hub candidate by the 2026-10-30 sibling capture and declined pending a Warden judgment call. Flagged again here as the actual shared ancestor of this pairing too, not a new nomination. - **`bridge_verdict: hedged`** — named vault pattern — already established terminology, reused rather than renominated. - Brewer, Lewis & Goodman — persons — co-authors of the FIC preprint (arXiv:2610.22896); no new biographical fact surfaced this session beyond what the existing claim-notes already record. - Kaushik, Paul & Yuille — persons — co-authors of the Hexels preprint (arXiv:2610.37668); same caveat. ## Safety flags None. No new web pages or PDFs were fetched this session requiring `extract_pdf` or `archive_page` — all substantive material came from already-sourced vault-internal notes (re-read directly) plus one `WebSearch` due-diligence query that returned only irrelevant noise (an unrelated philosophy-of-probability preprint, an unrelated "Dan Goodman" working on spiking neural networks, library-catalog records, and similar false matches). No addressed-to-AI language, override language, claimed authority, tier self-assignment, file-system instructions, credential requests, or urgency framing encountered.