LLMs cite less critically than humans, reshaping academic influence networks
Source published ·Modelwire updated
Original coverage: arXiv cs.CL ↗·How Modelwire adds context
The development
A new study reveals that LLMs cite scientific work with systematically different rhetorical patterns than human researchers, potentially reshaping how knowledge is attributed and valued in academic networks. Using a masked-citation task across six major models and 1,746 NLP papers, researchers found that language models tend to cite less critically, favoring neutral mentions over supportive or contrasting citations. This matters because citation behavior encodes social capital and intellectual influence. If LLM-assisted writing flattens the rhetorical texture of citations, it could subtly reshape which ideas gain traction and which researchers accumulate authority, with downstream effects on funding, hiring, and research direction.
Modelwire’s AI-generated summary of coverage from arXiv cs.CL.
Modelwire analysis
Analyst takeOur AI-generated reading of the wider context and the next developments to watch.
The study doesn't just show LLMs cite differently; it demonstrates that this difference systematically deprioritizes critical engagement. Models favor neutral mentions over supportive or contrasting citations, which means LLM-assisted papers may inadvertently flatten the rhetorical work that builds intellectual authority.
This connects directly to the August evaluation work on LLM-as-judge mechanisms (reference [1]) and human judgment alignment (reference [2]). Those papers exposed how LLMs make opaque, systematically biased decisions about quality and value. Citation behavior is the same problem in a different domain: LLMs are now encoding judgments about which ideas matter, which researchers matter, and which knowledge gets amplified. The difference is that citation bias operates at scale across the entire academic publishing pipeline, not just in isolated evaluation tasks. If LLM judges misalign with human reasoning about quality, LLM writers are now misaligning with human reasoning about intellectual contribution.
Monitor whether citation patterns in papers explicitly written with LLM assistance (disclosed in methods sections) show measurably flatter rhetorical texture than human-written controls within the same venues over the next 12-18 months. If the effect persists and grows as LLM adoption increases, watch for downstream shifts in citation counts and h-indices for researchers who adopt LLM writing tools versus those who don't.
This interpretation is generated from the summary above and the archive coverage cited below. Our methodology · Report an error
Coverage behind this analysis
These archive entries ground the connection in our analysis. They are ordered by source publication date, with links to our coverage and the original sources.
·arXiv cs.LG
Mechanistic analysis reveals how LLM judges evaluate text quality
Researchers have opened the black box of LLM-based evaluators by mechanistically analyzing how Themis and Prometheus assign quality ratings to generated text. Using controlled perturbations across readability and adequacy dimensions, combined with causal tracing and attention analysis, the work reveals that both models execute a coherent two-stage evaluation pipeline. This matters because LLM judges now…
MentionsLLMs · NLP · arXiv
How this coverage is produced
Modelwire uses AI to generate summaries and context from source headlines, snippets, and selected archive coverage. Automated checks do not verify every claim, and items are not routinely reviewed by a person before publication. Zacaria Solis operates the site. Read the linked source for the full evidence and report errors through our corrections process.
Modelwire summarizes, we don’t republish. arXiv cs.CL originally reported this story as “Citing Less Critically: LLMs Reshape the Rhetoric and Reach of Scientific Citation”. The full content lives on arxiv.org. If you’re a publisher and want a different summarization policy for your work, see our takedown page.