# Structurally Human, Semantically Biased: Detecting LLM-Generated References with Embeddings and GNNs

arXiv: https://arxiv.org/abs/2601.20704 · HTML: https://mlsanity.com/paper/2601.20704/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 29 of its 47 references that are in the map.
- First posted on arXiv 28 Jan 2026.

## Builds on (most cited first, 29 of 29)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Semi-Supervised Classification with Graph Convolutional Networks](https://mlsanity.com/paper/1609.02907/) (2016)
- [Graph Attention Networks](https://mlsanity.com/paper/1710.10903/) (2017)
- [Inductive Representation Learning on Large Graphs](https://mlsanity.com/paper/1706.02216/) (2017)
- [How Powerful are Graph Neural Networks?](https://mlsanity.com/paper/1810.00826/) (2018)
- [Neural Message Passing for Quantum Chemistry](https://mlsanity.com/paper/1704.01212/) (2017)
- [A Comprehensive Survey on Graph Neural Networks](https://mlsanity.com/paper/1901.00596/) (2019)
- [Graph Neural Networks: A Review of Methods and Applications](https://mlsanity.com/paper/1812.08434/) (2018)
- [A Survey on Evaluation of Large Language Models](https://mlsanity.com/paper/2307.03109/) (2023)
- [Pitfalls of Graph Neural Network Evaluation](https://mlsanity.com/paper/1811.05868/) (2018)
- [LLMs-as-Judges: A Comprehensive Survey on LLM-based Evaluation Methods](https://mlsanity.com/paper/2412.05579/) (2024)
- [A Fair Comparison of Graph Neural Networks for Graph Classification](https://mlsanity.com/paper/1912.09893/) (2019)
- [Language agents achieve superhuman synthesis of scientific knowledge](https://mlsanity.com/paper/2409.13740/) (2024)
- [SciRepEval: A Multi-Format Benchmark for Scientific Document Representations](https://mlsanity.com/paper/2211.13308/) (2022)
- [A Survey on Fairness in Large Language Models](https://mlsanity.com/paper/2308.10149/) (2023)
- [Large language models for automated scholarly paper review: A survey](https://mlsanity.com/paper/2501.10326/) (2025)
- [AI4Research: A Survey of Artificial Intelligence for Scientific Research](https://mlsanity.com/paper/2507.01903/) (2025)
- [Fairness in Large Language Models: A Taxonomic Survey](https://mlsanity.com/paper/2404.01349/) (2024)
- [On Positional and Structural Node Features for Graph Neural Networks on Non-attributed Graphs](https://mlsanity.com/paper/2107.01495/) (2021)
- [Automating Research Synthesis with Domain-Specific Large Language Model Fine-Tuning](https://mlsanity.com/paper/2404.08680/) (2024)
- [SciLitLLM: How to Adapt LLMs for Scientific Literature Understanding](https://mlsanity.com/paper/2408.15545/) (2024)
- [Automated scholarly paper review: Concepts, technologies, and challenges](https://mlsanity.com/paper/2111.07533/) (2021)
- [Large Language Models Reflect Human Citation Patterns with a Heightened Citation Bias](https://mlsanity.com/paper/2405.15739/) (2024)
- [How Deep Do Large Language Models Internalize Scientific Literature and Citation Practices?](https://mlsanity.com/paper/2504.02767/) (2025)
- [Who Gets Cited? Gender- and Majority-Bias in LLM-Driven Reference Selection](https://mlsanity.com/paper/2508.02740/) (2025)
- [Towards Better Generalization with Flexible Representation of Multi-Module Graph Neural Networks](https://mlsanity.com/paper/2209.06589/) (2022)
- [PaperAsk: A Benchmark for Reliability Evaluation of LLMs in Paper Search and Reading](https://mlsanity.com/paper/2510.22242/) (2025)
- [Who Gets Recommended? Investigating Gender, Race, and Country Disparities in Paper Recommendations from Large Language Models](https://mlsanity.com/paper/2501.00367/) (2024)
- [Generative AI and the future of scientometrics: current topics and future questions](https://mlsanity.com/paper/2507.00783/) (2025)

## Built on by (most cited first, 1 of 1)

- [Thinking Like a Scientist? A Structural Study of LLM-Generated Research Methods](https://mlsanity.com/paper/2606.26130/) (2026)

## Abstract (arXiv)

Large language models are increasingly used to curate bibliographies, raising the question: are their reference lists distinguishable from human ones? We build paired citation graphs, ground truth and GPT-4o-generated (from parametric knowledge), for 10,000 focal papers (≈ 275k references) from SciSciNet, and added a field-matched random baseline that preserves out-degree and field distributions while breaking latent structure. We compare (i) structure-only node features (degree/closeness/eigenvector centrality, clustering, edge count) with (ii) 3072-D title/abstract embeddings, using an RF on graph-level aggregates and Graph Neural Networks with node features. Structure alone barely separates GPT from ground truth (RF accuracy ≈ 0.60) despite cleanly rejecting the random baseline (≈ 0.89--0.92). By contrast, embeddings sharply increase separability: RF on aggregated embeddings reaches ≈ 0.83, and GNNs with embedding node features achieve 93% test accuracy on GPT vs. ground truth. We show the robustness of our findings by replicating the pipeline with Claude Sonnet 4.5 and with multiple embedding models (OpenAI and SPECTER), with RF separability for ground truth vs. Claude ≈ 0.77 and clean rejection of the random baseline. Thus, LLM bibliographies, generated purely from parametric knowledge, closely mimic human citation topology, but leave detectable semantic fingerprints; detection and debiasing should target content signals rather than global graph structure.