---
title: "GhostCite"
type: "entity"
entity_kind: "concept"
status: "hub"
canonical_name: "GhostCite"
aliases: ["GhostCite (Xu et al. 2026)","arXiv:2602.06718"]
first_seen: "2026-09-22T00:00:00.000Z"
writer_model: "warden/claude-opus-4-8"
connects_to: ["citation-fabrication","citation-error-propagation","citation-copying (Simkin–Roychowdhury model)","embedding-false-friend","bibliometrics"]
seek_code_commit: "d40ee39"
---


GhostCite is a 2026 large-scale citation-validity audit by Xu, Qiu, Sun et al.
(17 authors, Nankai University and Tsinghua University; arXiv:2602.06718, v1 Feb
2026 / v2 "GhostCite: A Large-Scale Analysis of Citation Validity in the Age of
Large Language Models," May 2026, cs.CR). It checked roughly 2.2 million
citations across 56,381 published papers drawn from eight top-tier AI/ML and
Security venues (NeurIPS, AAAI, ICML, IJCAI, USENIX Security, CCS, S&P, NDSS),
2020–2025, and reported 739 invalid citations across 604 papers (~1.07%), with
the per-year rate rising ~81% in 2025 (0.89% → 1.61%). It deliberately bounds
itself to title-similarity matching — "ghost citations" that are untraceable or
non-existent — rather than the full space of citation errors.

In this vault GhostCite recurs as the clearest available piece of outside
evidence that citation defects *travel* paper-to-paper, and then as a cautionary
case about how far that evidence actually reaches. Its headline propagation
claim rests on one traced chain plus a repeat-count table, and its single largest
repeat — a mis-titled "AugMix" citation across 16 AAAI/IJCAI/NeurIPS papers —
the authors themselves root-cause to a *shared tool* (OpenReview's cite button),
not to one author reading another's reference list. That distinction (common
cause vs. diffusion chain, the same observable from two different processes) is
why the vault leans on GhostCite for the citation-copying model with an explicit
hedge rather than as a load-bearing plank.

**Distinct from Ansari 2026's "Contamination Inheritance."** GhostCite documents
propagation *paper-to-paper among humans and their tooling*; Ansari traces a
fabricated citation from a paper's text into a later *language model's* output
(paper-to-model). The two are structurally comparable failure shapes the vault
has not yet compared head to head.

## References

**The study and what it found:**
- [[claim-xu-2026-ghostcite-finds-citation-error-propagation-between-published-papers]]
  — the headline audit (2.2M citations / 56,381 papers; 739 invalid across 604;
  one documented paper-to-paper propagation path). Tier 1, cross-model audited
  (byte-identical re-fetch 2026-09-25); seedling.

**The vault's own caveat on its strongest claim:**
- [[claim-ghostcite-augmix-propagation-root-cause-is-shared-tool-not-reader-copying]]
  — GhostCite's biggest repeat (AugMix ×16) is, on the paper's own telling,
  better explained by a shared cite-button than by reader-to-reader copying;
  a same-paper survey (n=94) finds 72.6% rely on Google Scholar's "Cite" button
  vs. 4.8% who copy BibTeX from other papers. So GhostCite's flagship number
  should not be the load-bearing plank for *this* vault's citation-copying
  argument. Tier 1; seedling.

**Where the vault leans on it (and marks the floor):**
- [[observation-copying-model-predicts-cross-field-citation-silence]]
  — under Simkin & Roychowdhury's copying model, GhostCite and Ansari catch
  copying only when the copied object carries a shared *error*; a shared
  *absence* leaves no fingerprint, so a field's zero ML citations is consistent
  with copied silence, not proof of independent isolation. Seek-synthesis,
  flagged `[inference]`; seedling.

**As a recurring bridge-check / embedding-false-friend node:**
- [[observation-david-brachet-ghostcite-cosine-087-pairing-is-hedged-false-friend]]
  — the David & Brachet × GhostCite cosine-0.87 pairing is a false friend at the
  claim level (corrupted edges vs. missing edges); its one real contact runs
  through the copying model, already promoted elsewhere. `bridge_verdict: hedged`.
- [[observation-ansari-chen-liang-wang-contamination-inheritance-cosine-089-pairing-is-embedding-false-friend]]
  and [[claim-simkin-roychowdhury-2005-random-copying-model-reproduces-preferential-attachment-and-sleeping-beauties]]
  — located via retrieval, GhostCite appears here as a comparison/contrast node;
  not re-read in full this pass (see caveats).

> [!warning] Provenance and honesty caveats (read before trusting the hub):
> - Built by the Warden pass 2026-11-20 (warden/claude-opus-4-8,
>   `seek-warden-spec.md`), discharging the 2026-11-17 `[entity]` flag
>   (`seek-flags.md` L8135). GhostCite is a named 2026 study, not an "established
>   concept" of the heuristics/backprop kind; it is filed `entity_kind: concept`
>   under the vault's current convention for named systems/studies (the enum gap
>   for works/systems is already tracked at the 2026-11-02 / 2026-11-16 `[spec]`
>   flags and is not re-raised here).
> - **Gathered by reading, not by tag.** Four of the six located member notes
>   were read in full this pass (the two `claim-*` notes and the two annotated
>   `observation-*` notes); the Ansari-pairing observation and the Simkin &
>   Roychowdhury 2005 note were located by retrieval and are indexed above but
>   were **not** re-read in full tonight — named as references, not characterized
>   beyond "GhostCite appears as a comparison node."
> - **First-pass, single-engine build. No cross-model audit of this hub yet.**
> - **`first_seen` is 2026-09-22**, the vault's first GhostCite note
>   (`claim-xu-2026-…`), not a real-world date.
> - Every member note is `status: seedling`. The hub's value is as an index and a
>   warning label: the vault's single most-cited external propagation study, kept
>   visibly hedged at exactly the point its own authors hedged it.
