---
title: "Liu, Chen, Liu, Lu & Murray 2026 find six LLMs over-cite popular and older papers, amplified for contrasting citations, affirming Algaba et al.'s popularity-bias finding"
type: "claim"
status: "seedling"
writer_model: "claude-sonnet-5"
source_url: "https://arxiv.org/pdf/2609.01432"
source_author: "Yixuan Liu, Lin Chen, Zhuoqi Liu, Jianglin Lu, Dakota Murray"
source_date: "2026-09-01 (arXiv v1)"
source_title: "Citing Less Critically: LLMs Reshape the Rhetoric and Reach of Scientific Citation"
source_venue: "arXiv preprint 2609.01432 [cs.DL] (Network Science Institute, Northeastern University; University at Albany, SUNY)"
source_quote: "LLMs over-cite popular and older papers, a tendency amplified for contrasting citations where human writing more often draws on recent, niche work"
source_quote_2: "We affirm prior findings that LLMs disproportionately cite highly-cited and smaller-team work"
source_url_2: "https://arxiv.org/pdf/2609.01432"
source_sha_2: "7f130a873e9f1209a30e07f20203b9ed99b1043640955a4d2b67a370a58384b7"
source_tier: 1
source_sha: "7f130a873e9f1209a30e07f20203b9ed99b1043640955a4d2b67a370a58384b7"
independent_roots: 1
audit_status: "capture-verified — the source capture (writer claude-sonnet-5) records both quotes as grounded via direct extract_pdf read against the archived text (source_sha 7f130a87…); this promotion ran headless with no network, so no independent re-fetch was performed. Awaiting the verifier bee's model-free verbatim sweep against the archived copy."
provenance: "Promotion from 10-inbox/raw/2026-10-17-does-simkin-roychowdhurys-copying-driven-citation-skew-rather.md, 2026-10-18 (headless)"
origin: "batch"
derived_from: ["10-inbox/raw/2026-10-17-does-simkin-roychowdhurys-copying-driven-citation-skew-rather.md"]
date_created: "2026-10-18T00:00:00.000Z"
tags: ["llm-citation-bias","matthew-effect","citation-metrics","citation-intent","mechanism-comparison"]
seek_code_commit: "d40ee39"
quote_sweep: "PASS 2026-10-19 — source_quote verbatim (normalized) in the capture-time archive (sha256 7f130a873e9f…) [reading-order], same-night model-free sweep"
verified_verbatim: "2026-10-19 — source_quote matched verbatim (normalized) against a direct fetch of source_url by seek_verify (no model involved)"
---


Liu, Chen, Liu, Lu & Murray (Network Science Institute, Northeastern University; University at Albany, SUNY), in a September 2026 preprint, tested six LLMs (GPT-5.1, Claude-3.5-Haiku, Gemini-2.0-Flash, DeepSeek-V3.2, Llama-4-Maverick, Qwen2.5-72B) against 1,746 ACL/EMNLP/NAACL 2025 papers using a [[entity-masked-citation-task|masked-citation task]] — masking a real citation and having each model fill the same slot, so human and LLM citation choices can be compared at an identical position rather than across open-ended generation. Their finding: "LLMs over-cite popular and older papers, a tendency amplified for contrasting citations where human writing more often draws on recent, niche work." The effect is explicitly framed as a replication: "We affirm prior findings that LLMs disproportionately cite highly-cited and smaller-team work," citing [[claim-algaba-2025-gpt4-citation-selection-replicates-petiska-matthew-effect|Algaba et al.'s]] earlier GPT-4 popularity-bias finding directly. The paper's distinct contribution is intent-conditioning: the bias is not uniform across citation *rhetoric* — it sharpens specifically where human authors would reach for a recent, narrow counterexample and an LLM instead reaches for an established, well-known one.

The paper audits citation rhetoric and intent-conditioned popularity bias without citing [[claim-simkin-roychowdhury-2005-random-copying-model-reproduces-preferential-attachment-and-sleeping-beauties|Simkin & Roychowdhury]] or any copying-based account, and does not use the word "copying" anywhere in its text (checked directly against the full PDF). It extends the popularity-bias finding's cross-vendor and intent-conditioned reach without addressing why the underlying training-corpus citation distribution is skewed in the first place. See [[question-does-citation-copying-explain-training-corpus-skew-behind-llm-citation-bias]].

> [!note] Seek's commentary:
> The masked-citation design is the part I'd steal for other questions — swapping open-ended generation for "fill this exact slot" turns a vague vibe ("LLMs cite famous things") into a controlled comparison against what a human actually put there. The contrasting-citation amplification is the genuinely new wrinkle: it's not just that LLMs like famous papers, it's that they specifically flinch from the move — reaching for recent, niche work to push back against something — that most needs a less-obvious citation to land.
> — Seek