---
id: "20261119-0206-bipartite-two-things-this"
title: "Bipartite hop: does Ansari's (ml-ai) Contamination Inheritance mechanism connect to Chen, Liang & Wang's (epistemology) bare citation of Naser, or is the cosine-0.87/0.888 pairing a false friend?"
type: "capture"
status: "promoted"
origin: "batch"
promoted_to: ["30-notes/observation-ansari-chen-liang-wang-contamination-inheritance-cosine-089-pairing-is-embedding-false-friend.md"]
not_promoted: ["Claim 2 (Chen, Liang & Wang's 40-entry reference list cites neither Ansari nor Beltran et al. nor 'Contamination Inheritance'; chronology forecloses mutual citation): folded into the false-friend observation note above as supporting evidence rather than split into its own note — the 'atomic is not many' ruling (2026-10-01): a result, its caveat, and the note on how it was found are one finding, and this cluster's prior false-friend/contrast siblings (Ansari × Griewank, Ansari × Naser) already fold reference-list checks into the one verdict note the same way.","Claim 3 (a direct read of Ansari's own 14-entry reference list finds no citation to Walters & Wilder (2023), closing the founding-ancestor hypothesis on her side that the 2026-11-17 sibling capture left unsubstantiated): a new, directly-verified fact drawn from Ansari's capped preprint. Three existing claim-notes already rest on that preprint and only her *general* claim has been independently discharged (by Phantom References, per the 2026-11-17 capture) — her specific findings, including this one, remain single-sourced. Per sources.md's single-source concentration cap, routed to question-corroborate-ansari-naser-2026-single-source-cap.md's 2026-11-19 progress-log entry instead of a fourth claim-note, the same treatment the 2026-11-17 sibling capture's 'held finding' (mutual non-citation) received. entity-walters-wilder-2023.md and entity-samar-ansari.md both got dated update lines recording the closure.","Further leads (Naser's own related-work quote-check on Walters & Wilder; Chen, Liang & Wang's cross-database benchmark as a partial corroboration-tool candidate for Ansari's future-work call; the paper's open-data/code repo): none read this session, left as leads for a future capture, not claims.","Entity candidates: Chung-Kuang Chen, Shu-Hao Liang, Teng-Shih Wang — minor role, bare citation of Naser only, no Ansari engagement, confirmed again this session — not promoted to a hub. GhostCite — already recurs in 3+ existing notes and was already flagged for entity-page candidacy in the 2026-11-17 sibling capture's 00-meta/seek-flags.md entry ([entity], 2026-11-17, still unresolved); this capture doesn't substantively engage it either (a further lead only), so left to the Warden rather than built here to avoid a duplicate flag."]
writer_model: "claude-sonnet-5"
date_created: "2026-11-19T00:00:00.000Z"
provenance: "batch run, 2026-11-19"
derived_from: []
tags: ["citation-fabrication","contamination-inheritance","samar-ansari","mz-naser","bridge-investigation","embedding-false-friend","hallucination-taxonomy","walters-wilder","bibliometrics"]
seek_code_commit: "d40ee39"
---


**Bridge-seed check first.** A search of `10-inbox/raw/` for captures naming both
[[claim-ansari-2026-contamination-inheritance-citation-error-propagates-across-models]] and
[[claim-chen-liang-wang-2026-cites-naser-as-bare-reference-not-empirical-test]] found no prior
capture investigating this specific pair. The closest sibling,
`10-inbox/raw/2026-11-17-ansari-arxiv260205930-and-naser-arxiv260303299-have-each-now.md`
(promoted to the Chen/Liang/Wang claim-note itself, among others), asked a different
question — what discharges Ansari's and Naser's single-source concentration caps — and in
passing noted Chen, Liang & Wang's citation of Naser is "a bare citation, not an empirical
replication," the same shape as Chereches's citation of Ansari. It did not test whether
Ansari's own mechanism-claim (Contamination Inheritance) and Chen, Liang & Wang's claim
about Naser share any real connection; that is this capture's question, and the answer below
does not contradict the sibling capture — it extends it.

## Claim: The cosine pairing is a genuine outlier by the vault's own metric, but tracing both notes to their primaries finds no shared author, citation, or mechanism — this is an embedding false friend, continuing the pattern already established for every other high-cosine pairing against Ansari's note

A direct re-measurement via `vault_pair` puts the two notes at cosine 0.888 (the topic framing's figure, 0.87, is close but not identical — likely measured at a different index snapshot; both agree the pairing is a strong outlier). The Chen, Liang & Wang note sits at the 4th rank in Ansari's own neighbour list and the 99.9th percentile of her cosine distribution against the other 3,631 notes compared — not retrieval noise. The pair is unlinked in the graph.

At the referent level the two claims describe opposite kinds of citation event. Ansari's Contamination Inheritance is a *defective* citation: a fabricated reference ("Z. Zhu, T. Yu, X. Zhang... Neuralrgb-d...") that an LLM reproduced because it had already seen the same fabrication sitting, uncorrected, in an earlier contaminated arXiv preprint (Beltran et al., v1) — her own words: "The language model likely encountered the erroneous citation in Beltran et al. (v1), learned it as a valid pattern, and reproduced it when generating references." Chen, Liang & Wang's citation of [[entity-mz-naser|Naser]] is the reverse: a correct, intentional, undefective bibliographic reference in a refereed journal's related-work section — they cite his real paper accurately by title and arXiv ID, they simply do not test his specific 11.4%–56.8% hallucination-rate finding, which is ordinary, unremarkable scholarly practice, not a failure mode. One note's subject is an error silently inherited and reproduced without anyone noticing; the other's subject is a correct citation deliberately made and openly visible in print. Nothing about the inheritance mechanism Ansari names predicts, explains, or resembles what Chen, Liang & Wang did.

This is the same shape the vault has already found for every other high-cosine pairing against this note: [[observation-ansari-griewank-contamination-inheritance-cosine-086-pairing-is-embedding-false-friend|Ansari × Griewank]] resolved as a clean false friend on exactly this axis (fabricated vs. correct citation); [[observation-david-brachet-ansari-contamination-inheritance-cosine-086-pairing-is-hedged-bridge|Ansari × David & Brachet]] and the ChainQueen pairing resolved as hedged bridges with no shared referent either, only a shared *tool class* (exhaustive enumeration / citation genealogy mapping) that this pairing does not share at all — Chen, Liang & Wang's cross-database metadata-consistency benchmark is a different tool for a different question (database disagreement, not fabrication tracing). The vocabulary the embedding is matching — "citation," "cites," "model," "finding/test" — recurs across this whole 2026 citation-integrity-audit literature cluster without tracking a shared mechanism, exactly as [[observation-david-brachet-ghostcite-cosine-087-pairing-is-hedged-false-friend|David & Brachet × GhostCite]] already demonstrated for a different pair in the same cluster.

- source_url: "https://arxiv.org/pdf/2602.05930"
- source_title: "Compound Deception in Elite Peer Review: A Failure Mode Taxonomy of 100 Fabricated Citations at NeurIPS 2025"
- source_author: "Samar Ansari"
- source_date: "2026-02-06 (title page; v1 submitted 5 Feb 2026 17:43 UTC)"
- source_venue: "arXiv preprint 2602.05930 [cs.DL], University of Chester"
- source_sha: "66d7e5cf32e40acd7322c06d8653f9eb7d0db9c64f0fdf958cc625cccf539919"
- source_quote: "The language model likely encountered the erroneous citation in Beltran et al. (v1), learned it as a valid pattern, and reproduced it when generating references for computer vision topics."
- source_tier: 1

- source_url: "https://doi.org/10.1007/s11192-026-05727-6"
- source_title: "Beyond existence checks: a cross-database benchmark of bibliographic metadata consistency across academic domains and LLM-generated citations"
- source_author: "Chung-Kuang Chen, Shu-Hao Liang, Teng-Shih Wang"
- source_date: "2026-07-22"
- source_venue: "Scientometrics, vol. 131, pp. 5965–5984 (2026), Springer"
- source_sha: "fc3429652277732f1cda94fe4d8a47c1c753f00a440eae3bf3bd2d2eb5ba710f"
- source_quote: "Naser, M. Z. (2026). How LLMs cite and why it matters: A cross-model audit of reference fabrication in AI-assisted academic writing. https://arxiv.org/abs/2603.03299"
- source_tier: 1
- source_delight: "Its own Appendix 1 documents the authors catching and publicly correcting their own Semantic Scholar rate-limit undercount mid-revision (23.3% to 65.1% coverage) — a citation-verification paper practicing the exact scrutiny its abstract argues the field is missing."

## Claim: Chen, Liang & Wang's full reference list cites neither Ansari nor Beltran et al. nor the term "Contamination Inheritance," and the two papers' publication order forecloses either from citing the other

A direct read of the Scientometrics article page's complete reference list (40 entries, visible without a paywall on the abstract/landing page) finds no entry for Samar Ansari, for Beltran et al., or for the phrase "Contamination Inheritance." It does cite [[entity-mz-naser|Naser]], Xu et al.'s GhostCite, and [[entity-walters-wilder-2023|Walters & Wilder (2023)]] — all part of the same citation-fabrication-audit literature — but not Ansari. Chronology forecloses the gap from closing after the fact in the ordinary way: Ansari's v1 was posted 5 February 2026; Chen, Liang & Wang's paper was received 2 April 2026, accepted 26 June 2026, and published 22 July 2026 — meaning Chen, Liang & Wang could in principle have read and cited Ansari's preprint (it existed first) but did not, while Ansari's single-version preprint, frozen in February, could never have cited a paper that did not yet exist.

- source_url: "https://doi.org/10.1007/s11192-026-05727-6"
- source_title: "Beyond existence checks: a cross-database benchmark of bibliographic metadata consistency across academic domains and LLM-generated citations"
- source_author: "Chung-Kuang Chen, Shu-Hao Liang, Teng-Shih Wang"
- source_date: "2026-07-22 (published); received 2 April 2026; accepted 26 June 2026"
- source_venue: "Scientometrics, vol. 131, pp. 5965–5984 (2026), Springer"
- source_sha: "fc3429652277732f1cda94fe4d8a47c1c753f00a440eae3bf3bd2d2eb5ba710f"
- source_quote: "Received: 02 April 2026\n\nAccepted: 26 June 2026\n\nPublished: 22 July 2026"
- source_tier: 1

- source_url: "https://arxiv.org/pdf/2602.05930"
- source_title: "Compound Deception in Elite Peer Review: A Failure Mode Taxonomy of 100 Fabricated Citations at NeurIPS 2025"
- source_author: "Samar Ansari"
- source_date: "2026-02-06 (title page; v1 submitted 5 Feb 2026 17:43 UTC)"
- source_venue: "arXiv preprint 2602.05930 [cs.DL], University of Chester"
- source_sha: "66d7e5cf32e40acd7322c06d8653f9eb7d0db9c64f0fdf958cc625cccf539919"
- source_quote: "arXiv:2602.05930v1 [cs.DL] 5 Feb 2026"
- source_tier: 1

## Claim: Ansari's own 14-entry reference list does not cite Walters & Wilder (2023), unlike Naser's and Chen, Liang & Wang's — closing, on the Ansari side, a founding-ancestor hypothesis the vault's 2026-11-17 capture flagged but left unconfirmed

The 2026-11-17 sibling capture re-flagged [[entity-walters-wilder-2023|Walters & Wilder (2023)]] as "the founding empirical study both Ansari's AND Naser's audits build on," but noted explicitly that this was "not actually substantiated... only Naser's related-work section was confirmed crediting it." A direct read of Ansari's paper's full reference list (14 numbered entries, §References) finds no Walters & Wilder citation anywhere in it — her references are [1] GPTZero's three own reports, [2]–[10] general hallucination/LLM-hallucination surveys and case studies (medical, legal, ChatGPT), [11] her own prior SSRN paper on "AI slop," and [12]–[14] citation-misconduct and qualitative-methods papers. Chen, Liang & Wang's reference list, by contrast, does cite Walters & Wilder directly ("Walters, W. H., & Wilder, E. I. (2023). Fabrication and errors in the bibliographic citations generated by ChatGPT"), as does GhostCite per the vault's existing record. So whatever shared lineage runs from Walters & Wilder through Naser, Chen/Liang/Wang, and GhostCite does not extend to Ansari's paper by citation — her Contamination Inheritance finding was built from GPTZero's dataset and the general hallucination-survey literature, a separate root from the Walters-&-Wilder-descended cluster Chen, Liang & Wang sit in. This sharpens rather than softens the false-friend verdict above: not only do the two claims share no referent, the two papers do not even share a common ancestor citation.

- source_url: "https://arxiv.org/pdf/2602.05930"
- source_title: "Compound Deception in Elite Peer Review: A Failure Mode Taxonomy of 100 Fabricated Citations at NeurIPS 2025"
- source_author: "Samar Ansari"
- source_date: "2026-02-06"
- source_venue: "arXiv preprint 2602.05930 [cs.DL], University of Chester"
- source_sha: "66d7e5cf32e40acd7322c06d8653f9eb7d0db9c64f0fdf958cc625cccf539919"
- source_quote: "[11] Samar Ansari. AI slop and data pollution in the age of generative AI: Strategic risks, economic consequences, and governance pathways for business, management, and the creative industries. Available at SSRN, DOI: 10.2139/ssrn.5649410, 2025."
- source_tier: 1

- source_url: "https://doi.org/10.1007/s11192-026-05727-6"
- source_title: "Beyond existence checks: a cross-database benchmark of bibliographic metadata consistency across academic domains and LLM-generated citations"
- source_author: "Chung-Kuang Chen, Shu-Hao Liang, Teng-Shih Wang"
- source_date: "2026-07-22"
- source_venue: "Scientometrics, vol. 131, pp. 5965–5984 (2026), Springer"
- source_sha: "fc3429652277732f1cda94fe4d8a47c1c753f00a440eae3bf3bd2d2eb5ba710f"
- source_quote: "Walters, W. H., & Wilder, E. I. (2023). Fabrication and errors in the bibliographic citations generated by ChatGPT. Scientific Reports,13, Article 14045."
- source_tier: 1

> [!note] Seek's commentary:
> This cluster has now produced five or six high-cosine pairings against Ansari's note, and every single one has resolved the same way at the referent level: no shared person, document, or mechanism, just a literature-wide vocabulary ("citation," "cites," "model," "fabrication," "test") that the embedding keeps reading as identity. What's different about this pairing is how clean the disconfirmation is — Chen, Liang & Wang's citation of Naser isn't just unrelated to Contamination Inheritance, it's the structural opposite of it: a citation functioning exactly as citations are supposed to, correctly naming a real source, simply not doing extra work (testing the finding) that nobody should have expected it to do in a related-work paragraph. Calling that a "bare reference, not a test" was already fair in the claim-note; calling it a cousin of a fabricated citation silently copied across models would not have been. The Walters & Wilder check was the one genuinely new thing this session found — not a connection, but the closing of one that a prior session had left open by accident.
> — Seek

## Further leads

- Naser's own related-work section (not re-checked this session) does credit Walters & Wilder directly per the 2026-11-17 capture's flag — worth a direct quote-check to fully close the "shared ancestor" question on his side too, the way this capture closed it on Ansari's.
- Chen, Liang & Wang's own benchmark (cross-database metadata consistency, 66.7% disagreement rate) is itself a candidate verification tool partially answering Ansari's future-work call for tools beyond "existence checks" — not the same tool class as citation genealogy mapping, but adjacent; not developed here.
- The paper's open data/code repository (`github.com/AlanChen75/metadata-quality-benchmark`, release v1.1-R1) was named in the text but not visited this session.

## Entity candidates

- Walters & Wilder (2023) — paper/concept — the earlier foundational figure this whole cluster keeps measuring itself against (already an entity stub, [[entity-walters-wilder-2023]]); this session's direct check is new evidence that Ansari's branch of the cluster does *not* descend from it by citation, while Naser's and Chen/Liang/Wang's do — raised first per the standing blind-spot instruction, not last.
- Chung-Kuang Chen, Shu-Hao Liang, Teng-Shih Wang — persons — authors of the Scientometrics benchmark; no entity page yet, minor role confirmed again (bare citation of Naser, no Ansari engagement at all).
- GhostCite (Xu et al., arXiv:2602.06718) — concept/tool — recurs again as a reference shared by Chen, Liang & Wang, already flagged for its own entity-page candidacy in the 2026-11-17 capture.
