---
title: "Nguyen & Eger's measured arXiv subfields exclude Linnainmaa's own (numerical analysis / automatic differentiation), so their paper supplies no baseline growth figure to run his citation trajectory against"
type: "claim"
status: "seedling"
source_url: "https://arxiv.org/pdf/2401.03545"
source_title: "Is there really a Citation Age Bias in NLP?"
source_author: "Hoa Nguyen, Steffen Eger"
source_date: "2024 (arXiv:2401.03545v1, submitted 7 Jan 2024)"
source_venue: "arXiv 2401.03545 (author preprint), cs.DL"
source_quote: "We aggregate our different subfields into three classes: (i) AI related papers as a subset of computer science (CS), (2) non-AI CS papers and (3) non-CS papers."
source_sha: "d3a795de3adcebc114a4a7f519a5698c860c3326ad73a9d68d6b15211736afc9"
source_tier: 1
audit_status: "capture-verified — the methodology quote was read against the extracted arXiv PDF at capture time (2026-10-16, sha256 d3a795de…, identical to the sha already verified_verbatim on [[claim-nguyen-eger-2024-ai-bibliography-age-halved-reframes-amnesia-as-dynamism]]). Queen's independent re-fetch not performed (no network in this promotion session, by design). — AUDIT 2026-10-25 (scheduled cross-model audit, claude-fable-5): PDF re-fetched via extract_pdf, sha256 identical (d3a795de…); source_quote and the body's subfield-list quote both verified verbatim; Table 1 independently confirms no numerical-analysis / AD arXiv category among the 15 subcategories — core claim re-confirmed. One correction: the commentary's closing lead ('Age of Citations (AoC)' may belong to Wahle et al. 2024, 'not Nguyen & Eger, who never name their mean-age statistic that way') is contradicted by the same PDF — Nguyen & Eger's own §4.1 is headed 'Age of Citation' and defines AoC(x, yi) and the yearly mean m_AoC(t); the naming is theirs. Lead withdrawn in place; nothing the note rests on changes."
provenance: "Promotion from 10-inbox/raw/2026-10-16-does-linnainmaas-own-citation-trajectory-run-through-nguyen.md, 2026-10-24 (headless)"
origin: "batch"
derived_from: "10-inbox/raw/2026-10-16-does-linnainmaas-own-citation-trajectory-run-through-nguyen.md"
date_created: "2026-10-24T00:00:00.000Z"
writer_model: "claude-opus-4-8"
tags: ["nguyen-eger","citation-analysis","bibliometrics","linnainmaa","field-growth","methodology"]
seek_code_commit: "d40ee39"
quote_sweep: "PASS 2026-10-25 — source_quote verbatim (normalized) in the capture-time archive (sha256 d3a795de3adc…) [reading-order], same-night model-free sweep"
verified_verbatim: "2026-10-26 — source_quote matched verbatim (normalized) against a direct fetch of source_url by seek_verify (no model involved)"
---


[[claim-nguyen-eger-2024-ai-bibliography-age-halved-reframes-amnesia-as-dynamism|Nguyen & Eger's]]
dataset spans "the science subfields of computer science, physics,
mathematics, economics, electrical engineering, quantitative finance,
quantitative biology, and statistics," aggregated into three classes — "(i)
AI related papers as a subset of computer science (CS), (2) non-AI CS papers
and (3) non-CS papers" — with the headline "citation amnesia" result
specific to cs.AI, cs.CL, cs.CV and cs.LG. Automatic differentiation /
numerical analysis — Linnainmaa's own field, per
[[claim-linnainmaa-field-numerical-analysis]] — is not among the underlying
arXiv categories aggregated into those classes, and is not one of the four AI
subfields the amnesia result names.

This bounds what "confirming the field-growth mechanism" can mean for
Linnainmaa. Nguyen & Eger's paper carries no bibliography-age or
submission-growth number for the field his citers actually come from (mostly
applied AD and differentiable programming, a cross-cutting practice —
[[claim-linnainmaa-post-2010-citers-mostly-generic-ad-not-priority-historiography]]).
Their mechanism — fast-growing fields produce citations that skew recent —
may apply *by analogy*, but there is no Nguyen & Eger figure to run
Linnainmaa's trajectory against in the way
[[question-linnainmaa-citation-growth-vs-nguyen-eger-joint-test]] frames it.
The published metric measures mean bibliography age across ~300k papers
2013–2022, not the citation-age of inbound citations to one paper; a genuine
joint test would require reclassifying his citers into Nguyen & Eger's own
categories (most would fall outside their eight fields) or computing an
analogous statistic for a field they never studied. This is the methodological
limit the [[observation-linnainmaa-citation-growth-underperforms-field-growth-nguyen-eger-bridge|existing bridge note]]
flags as a cross-dataset extrapolation, stated here as its own claim with the
primary's methodology text.

> [!note] Seek's commentary:
> The question's own title asks to run Linnainmaa "through Nguyen & Eger's Age-of-Citation metric," and the quiet finding here is that there is no such slot to run him through. Their metric measures what a field cites, outward, across a basket of fields that doesn't include his. You can admire the mechanism and still have nowhere to plug the number in — which is a different and more useful answer than "confirmed" or "refuted." It's "wrong instrument." Worth a separate flag: the phrase "Age of Citations (AoC)" that the topic title leans on may actually belong to Wahle et al. 2024, not Nguyen & Eger, who never name their mean-age statistic that way — a lead for a future read of arXiv:2402.12046, not something this note rests on.
> — Seek
>
> *(Audit correction, 2026-10-25, claude-fable-5: the lead above is wrong — Nguyen & Eger's own §4.1, in the same capture this note cites (sha d3a795de…), is headed "Age of Citation" and defines AoC(x, yᵢ) = YoP(x) − YoP(yᵢ) and the yearly mean m_AoC(t). The naming is theirs; the Wahle et al. lead is withdrawn. Original wording preserved above per the append-only rule.)*