---
id: "20260708-0208-answered-2026-07-07"
title: "Capture: Dedupe-check on '[ANSWERED 2026-07-07, cycle 17] Griewank-2014 Table 5' topic — confirms fold-in is real, no new research performed"
type: "capture"
status: "promoted"
promoted_to: []
not_promoted: ["Griewank-2014 Table 5 fold-in — already load-bearing in [[claim-linnainmaa-reverse-mode-single-pass]] (Revisit cycle 17) and [[claim-reverse-mode-multiple-independent-discovery]] (Ostrowski receipt). Dedupe confirmation only; nothing new to promote."]
promotion_note: "Processed 2026-07-09 (Fable clean-lane promotion marathon): verified both named fold-ins exist and are accurate; no new notes. Capture flags a queue bug — already-[ANSWERED] topics are being re-queued into the bee (harvest/schedule step not honoring the [ANSWERED] prefix)."
origin: "batch"
model: "claude-sonnet-5"
date_created: "2026-07-08T00:00:00.000Z"
provenance: "batch run 2026-07-08, cycle N+1; topic string arrived already marked '[ANSWERED 2026-07-07, cycle 17] Griewank-2014 Table 5 → yes; folded into claim-linnainmaa-reverse-mode-single-pass + Ostrowski receipt into claim-reverse-mode-multiple-independent-discovery', itself harvested from 10-inbox/raw/2026-06-28-read-griewank-2012-directly-to-verify-a-the-backward-pass-structural-description-of-linnainmaas-algorithm-and-b-the-copenhagen-park-anecdote.md"
derived_from: []
tags: ["automatic-differentiation","linnainmaa","griewank","reverse-mode","history-of-ml","dedupe","housekeeping"]
source_url: "https://www.scielo.br/j/pope/a/kZ8M4LPYFryXQkdwHz7QXVx/?format=pdf&lang=en"
source_author: "Andreas Griewank"
source_date: "2014"
source_tier: 1
---


> [!info] Outcome: no new research — this is a duplicate of already-completed work
> This topic string, as received, already declares its own resolution: `[ANSWERED 2026-07-07, cycle 17] Griewank-2014 Table 5 → yes; folded into claim-linnainmaa-reverse-mode-single-pass + Ostrowski receipt into claim-reverse-mode-multiple-independent-discovery`. Rather than re-running web research that had already reached Tier 1 primary-source depth, this capture verifies the fold-in claim against the current state of the vault and confirms it is accurate. No web fetches were performed beyond checking that the prior primary source is still the one on record.

## What was checked

The topic asserts two fold-in destinations. Both were opened and read directly (not assumed from the topic string):

1. **`30-notes/claim-linnainmaa-reverse-mode-single-pass.md`** — contains a "Revisit 2026-07-07 (queen cycle 17)" section citing Griewank 2014's Table 5 ("Adjoint recursion for general evaluation procedure") and the explicit attribution to Linnainmaa's 1970 thesis, matching the topic's claim. The `verified_by` frontmatter field also names the exact capture (`10-inbox/raw/2026-07-01-does-table-5-adjoint-recursion-griewank-2014-support-tier-1-linnainmaa-backward-pass.md`) as the source of this verification.
2. **`30-notes/claim-reverse-mode-multiple-independent-discovery.md`** — contains an "Update 2026-07-07 (queen cycle 17) — Ostrowski receipt upgraded" section that cites the same 2026-07-01 capture for a Tier-1 bibliographic entry: "OSTROVSKII GM, VOLIN YU M & BORISOV WW. 1971. Über die Berechnung von Ableitungen. Wiss. Z. Tech. Hochschule für Chemie, 13: 382–384." — including an honest discrepancy flag that the 1971 publication date postdates Linnainmaa's 1970 thesis, distinct from Griewank's claim that Ostrowski's group *used* the technique ~1965 (a use-date vs. publication-date distinction the note preserves rather than collapsing).

Both fold-ins trace to the same underlying Tier-1 source already on file in the vault:

**Claim type**: technical-mechanism / historical (load-bearing attribution)
**Sourcing floor**: Tier 1-2 required — met.

| Field | Value |
|---|---|
| source_title | "On Automatic Differentiation and Algorithmic Linearization" |
| source_author | Andreas Griewank |
| source_date | 2014 |
| source_venue | Pesquisa Operacional 34(3): 621-645, Brazilian Operations Research Society |
| source_url | https://www.scielo.br/j/pope/a/kZ8M4LPYFryXQkdwHz7QXVx/?lang=en |
| source_pdf_url | https://www.scielo.br/j/pope/a/kZ8M4LPYFryXQkdwHz7QXVx/?format=pdf&lang=en |
| source_tier | 1 |
| exact_quote (Table 5 attribution) | "Apparently the first author to write down this reverse procedure was Seppo Linnainmaa, who listed it in Fortran at the end of his Master Thesis [23], which is otherwise written in Finnish." (p. 629) |
| exact_quote (Ostrowski bibliographic receipt) | "OSTROVSKII GM, VOLIN YU M & BORISOV WW. 1971. Über die Berechnung von Ableitungen. Wiss. Z. Tech. Hochschule für Chemie, 13: 382–384." (Griewank 2014 bibliography, p. 645) |

Both notes' `status` remains `budding` and neither required a correction as part of this dedupe check — the fold-in described in the topic string is accurate and already load-bearing in the vault, not merely asserted.

## Why no further web research was performed

The rubric's sourcing floor (Tier 1-2 for technical-mechanism and quantitative/historical load-bearing claims) was already met by a direct PDF read of the Griewank 2014 primary source in the 2026-07-01 capture, and cross-checked again during 2026-07-06 audit remediation per the `audit_status: "verified-verbatim"` field on `claim-linnainmaa-reverse-mode-single-pass.md`. Re-fetching the same PDF a third time would not add new evidence; it would only re-confirm what two prior passes already confirmed at the ceiling tier for this claim type. Per the operating spec's instruction that a complete, non-forced outcome is valid, this capture records the dedupe finding rather than manufacturing new research on an already-closed question.

> [!note] Seek's commentary:
> This looks like a queue/scheduling artifact rather than a live research gap — the topic arrived pre-labeled as answered, and checking the two named destination notes confirms the label is honest, not just asserted by whoever wrote the topic string. Worth flagging upstream (outside this capture's scope) that the batch queue is re-surfacing topics that already carry an `[ANSWERED ...]` prefix in cycle 17; if that prefix is meant to suppress re-queuing, something in the harvesting/scheduling step isn't reading it.