---
title: "Monitoring AI-Modified Content at Scale: A Case Study on the"
description: "Their framework, distributional GPT quantification, builds word-frequency distributions from reference… 87 references, cited by 50 papers here."
authors: ["Weixin Liang", "Zachary Izzo", "Yaohui Zhang", "Haley Lepp", "Hancheng Cao", "Xuandong Zhao", "Lingjiao Chen", "Haotian Ye", "Sheng Liu", "Zhi Huang", "Daniel A. McFarland", "James Y. Zou"]
year: 2024
venue: "arXiv:2403.07183"
source: "https://arxiv.org/abs/2403.07183"
doi: "10.48550/arxiv.2403.07183"
arxiv: "2403.07183"
licence: "arXiv non-exclusive distribution license"
licence_url: "http://arxiv.org/licenses/nonexclusive-distrib/1.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/liang-2024-monitoring-ai-modified-content/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/liang-2024-monitoring-ai-modified-content/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2024](https://md.research.elamur.ai/free/en/papers/2024/page/3/) / Monitoring AI-Modified Content at Scale: A Case Study on the Impact of ChatGPT on AI Conference Peer Reviews

# Monitoring AI-Modified Content at Scale: A Case Study on the Impact of ChatGPT on AI Conference Peer Reviews

**Authors:** [Weixin Liang](https://md.research.elamur.ai/free/en/authors/weixin-liang/), Zachary Izzo, [Yaohui Zhang](https://md.research.elamur.ai/free/en/authors/yaohui-zhang/), [Haley Lepp](https://md.research.elamur.ai/free/en/authors/haley-lepp/), [Hancheng Cao](https://md.research.elamur.ai/free/en/authors/hancheng-cao/), [Xuandong Zhao](https://md.research.elamur.ai/free/en/authors/xuandong-zhao/), [Lingjiao Chen](https://md.research.elamur.ai/free/en/authors/lingjiao-chen/), [Haotian Ye](https://md.research.elamur.ai/free/en/authors/haotian-ye/), [Sheng Liu](https://md.research.elamur.ai/free/en/authors/sheng-liu/), [Zhi Huang](https://md.research.elamur.ai/free/en/authors/zhi-huang/), [Daniel A. McFarland](https://md.research.elamur.ai/free/en/authors/daniel-a-mcfarland/) and [James Y. Zou](https://md.research.elamur.ai/free/en/authors/james-zou/) · **Year:** 2024 · **Venue:** arXiv:2403.07183 · **Licence:** [arXiv non-exclusive distribution license](http://arxiv.org/licenses/nonexclusive-distrib/1.0/)

**Links:** [Publisher's page](https://arxiv.org/abs/2403.07183) · [PDF](https://arxiv.org/pdf/2403.07183) · [DOI 10.48550/arxiv.2403.07183](https://doi.org/10.48550/arxiv.2403.07183) · [arXiv 2403.07183](https://arxiv.org/abs/2403.07183) · [OpenAlex](https://openalex.org/W4392781833)

**Citations:** 69 citations (OpenAlex, 2026-10-14) · [50](https://md.research.elamur.ai/free/en/papers/liang-2024-monitoring-ai-modified-content/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
Detecting LLM-written text one document at a time is unreliable, and humans perform barely better than chance at it. So the scale of real-world LLM use has been hard to measure. The authors therefore estimate usage for a whole corpus rather than for single texts. Their framework, distributional GPT quantification, builds word-frequency distributions from reference texts known to be written by human experts or generated by AI. It then applies maximum likelihood estimation to infer what share of a target corpus is AI-generated or heavily AI-edited, without classifying any individual review.

They applied the method to peer reviews from ICLR 2024, NeurIPS 2023, CoRL 2023 and EMNLP 2023, and used Nature portfolio journal reviews as a control. Between 6.5% and 16.9% of review text at these AI venues appears substantially modified by LLMs, beyond simple grammar fixes. The Nature reviews show no such signal. Higher estimated LLM use coincides with:
- low reviewer confidence
- submission close to the deadline
- lower engagement with author rebuttals

The approach offers a cheap, scalable way to monitor AI influence on information ecosystems.

## Key points
- **Venue estimates:** About 10.6% of ICLR 2024 review sentences and 16.9% of EMNLP 2023 review sentences are estimated to be substantially LLM-modified. Nature portfolio reviews show no significant evidence of ChatGPT use.
- **Efficiency and accuracy:** The estimator is roughly 10 million times (7 orders of magnitude) cheaper computationally than state-of-the-art instance-level detectors. It also cuts estimation error by a factor of 3.4 in-distribution and 4.6 out-of-distribution.
- **Adjective signal:** Certain adjectives jumped sharply in ICLR 2024 reviews, including "commendable" (9.8×), "meticulous" (34.7×) and "intricate" (11.2×). Adjectives serve as the main features, although adverbs, verbs and non-technical nouns give similar results.
- **Behavioural correlates:**
  - Estimated AI use rises within 3 days of review deadlines.
  - It is higher among reviewers who skip ICLR/NeurIPS author rebuttals.
  - It is lower in reviews that cite scholarly work.
  - It is linked to more homogeneous review content in embedding space.
- **Robustness and availability:** A model trained only on GPT-3.5 data also detects GPT-4 text accurately. The paper appeared at ICML 2024 (PMLR 235), and the code is available on GitHub (Weixin-Liang/Mapping-the-Increasing-Use-of-LLMs-in-Scientific-Papers).

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

## Institutions

- [Stanford University](https://md.research.elamur.ai/free/en/universities/stanford-university/) — University, Stanford, United States · 506 works · 330,053 citations
- [University of California, Santa Barbara](https://md.research.elamur.ai/free/en/universities/university-of-california-santa-barbara/) — University, Santa Barbara, United States · 67 works · 10,550 citations
- [NEC (United States)](https://md.research.elamur.ai/free/en/companies/nec/) — Company, Irving, United States · 10 works · 658 citations

📚 Shelves (6): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 18,614 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 17,965 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 17,450 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 17,296 works · [🤝 Social Sciences](https://md.research.elamur.ai/free/en/shelves/social-sciences/) · 1,954 works · [🤝 Safety Research](https://md.research.elamur.ai/free/en/shelves/social-sciences/safety-research/) · 626 works

## Cite

Cite as: Weixin Liang, Zachary Izzo, Yaohui Zhang, Haley Lepp, Hancheng Cao, Xuandong Zhao, Lingjiao Chen, Haotian Ye, Sheng Liu, Zhi Huang, Daniel A. McFarland and James Y. Zou. 2024. Monitoring AI-Modified Content at Scale: A Case Study on the Impact of ChatGPT on AI Conference Peer Reviews. arXiv:2403.07183.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@misc{liang2024monitoring,
  title = {{Monitoring AI-Modified Content at Scale: A Case Study on the Impact of ChatGPT on AI Conference Peer Reviews}},
  author = {Weixin Liang, Zachary Izzo, Yaohui Zhang, Haley Lepp, Hancheng Cao, Xuandong Zhao, Lingjiao Chen, Haotian Ye, Sheng Liu, Zhi Huang, Daniel A. McFarland and James Y. Zou},
  year = {2024},
  doi = {10.48550/arxiv.2403.07183},
  eprint = {2403.07183},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2403.07183}
}
```

- [← previous Llumnix: Dynamic Scheduling for Large Language Model Serving](https://md.research.elamur.ai/free/en/papers/sun-2024-llumnix-dynamic-scheduling-large-language/)
- [next → SALAD-Bench: A Hierarchical and Comprehensive Safety Benchmark for Large Language Models](https://md.research.elamur.ai/free/en/papers/li-2024-salad-bench-hierarchical-comprehensive-safety/)
- [↑ back to papers of 2024](https://md.research.elamur.ai/free/en/papers/2024/page/3/)

## References

87 references, 31 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Arora, A., Kaffee, L.-A., and Augenstein, I. [Probing pre-trained language models for cross-cultural differences in values](https://md.research.elamur.ai/free/en/papers/arora-2023-probing-pre-trained-language-models/). arXiv preprint arXiv:2203.13722, 2022.
2. Atallah, M. J., Raskin, V., Crogan, M., Hempelmann, C. F., Kerschbaum, F., Mohamed, D., and Naik, S. Natural Language Watermarking: Design, Analysis, and a Proof-of-Concept Implementation. In Information Hiding, 2001. [find](https://scholar.google.com/scholar?q=Atallah%2C+M%2E+J%2E%2C+Raskin%2C+V%2E%2C+Crogan%2C+M%2E%2C+Hempelmann%2C+C%2E+F%2E%2C+Kerschbaum%2C+F%2E%2C+Mohamed%2C+D%2E%2C+and+Naik%2C+S%2E+Natural+Language+Watermarking%3A+Design%2C+Analysis%2C+and+a+Proof%2Dof%2DConcept+Implementation%2E+In+Informat)
3. Atallah, M. J., Raskin, V., Hempelmann, C. F., Topkara, M., Sion, R., Topkara, U., and Triezenberg, K. E. Natural Language Watermarking and Tamperproofing. In Information Hiding, 2002. [find](https://scholar.google.com/scholar?q=Atallah%2C+M%2E+J%2E%2C+Raskin%2C+V%2E%2C+Hempelmann%2C+C%2E+F%2E%2C+Topkara%2C+M%2E%2C+Sion%2C+R%2E%2C+Topkara%2C+U%2E%2C+and+Triezenberg%2C+K%2E+E%2E+Natural+Language+Watermarking+and+Tamperproofing%2E+In+Information+Hiding%2C+2002%2E)
4. Badaskar, S., Agarwal, S., and Arora, S. Identifying Real or Fake Articles: Towards better Language Modeling. In International Joint Conference on Natural Language Processing, 2008. URL [https://api.semanticscholar.org/CorpusID:4324753](https://api.semanticscholar.org/CorpusID:4324753). [↗](https://api.semanticscholar.org/CorpusID:4324753)
5. Bakhtin, A., Gross, S., Ott, M., Deng, Y., Ranzato, M., and Szlam, A. Real or Fake? Learning to Discriminate Machine from Human Generated Text. ArXiv, abs/1906.03351, 2019. URL [https://api.semanticscholar.org/CorpusID:](https://api.semanticscholar.org/CorpusID:) 182952342. — [arXiv:1906.03351](https://arxiv.org/abs/1906.03351) [↗](https://api.semanticscholar.org/CorpusID:)
6. Bao, G., Zhao, Y., Teng, Z., Yang, L., and Zhang, Y. [Fast-DetectGPT: Efficient Zero-Shot Detection of Machine-Generated Text via Conditional Probability Curvature](https://md.research.elamur.ai/free/en/papers/bao-2023-fast-detectgpt-efficient-zero-shot/). ArXiv, abs/2310.05130, 2023. URL [https://api.semanticscholar.org/CorpusID:263831345](https://api.semanticscholar.org/CorpusID:263831345). [↗](https://api.semanticscholar.org/CorpusID:263831345)
7. Bearman, M., Ryan, J., and Ajjawi, R. Discourses of artificial intelligence in higher education: A critical literature review. Higher Education, 86(2):369–385, 2023. [find](https://scholar.google.com/scholar?q=Bearman%2C+M%2E%2C+Ryan%2C+J%2E%2C+and+Ajjawi%2C+R%2E+Discourses+of+artificial+intelligence+in+higher+education%3A+A+critical+literature+review%2E+Higher+Education%2C+86%282%29%3A369%E2%80%93385%2C+2023%2E)
8. Beresneva, D. Computer-Generated Text Detection Using Machine Learning: A Systematic Review. In International Conference on Applications of Natural Language to Data Bases, 2016. URL [https://api.semanticscholar.org/CorpusID:1175726](https://api.semanticscholar.org/CorpusID:1175726). [↗](https://api.semanticscholar.org/CorpusID:1175726)
9. Bhagat, R. and Hovy, E. H. [Squibs: What Is a Paraphrase?](https://md.research.elamur.ai/free/en/papers/bhagat-2013-squibs-what-is-paraphrase/) Computational Linguistics, 39:463–472, 2013. URL [https://api.semanticscholar.org/CorpusID:32452685](https://api.semanticscholar.org/CorpusID:32452685). [↗](https://api.semanticscholar.org/CorpusID:32452685)
10. Bhattacharjee, A., Kumarage, T., Moraffah, R., and Liu, H. ConDA: Contrastive Domain Adaptation for AI-generated Text Detection. ArXiv, abs/2309.03992, 2023. URL [https://api.semanticscholar.org/CorpusID:261660497](https://api.semanticscholar.org/CorpusID:261660497). [↗](https://api.semanticscholar.org/CorpusID:261660497)
11. Bommasani, R., Creel, K. A., Kumar, A., Jurafsky, D., and Liang, P. S. Picking on the same person: Does algorithmic monoculture lead to outcome homogenization? Advances in Neural Information Processing Systems, 35: 3663–3678, 2022. [find](https://scholar.google.com/scholar?q=Bommasani%2C+R%2E%2C+Creel%2C+K%2E+A%2E%2C+Kumar%2C+A%2E%2C+Jurafsky%2C+D%2E%2C+and+Liang%2C+P%2E+S%2E+Picking+on+the+same+person%3A+Does+algorithmic+monoculture+lead+to+outcome+homogenization%3F+Advances+in+Neural+Information+Processin)
12. Cantor, M. Nearly 50 news websites are ‘AI-generated’, a study says. Would I be able to tell?, 2023. URL [https://www.theguardian.com/technology/2023/may/](https://www.theguardian.com/technology/2023/may/) 08/ai-generated-news-websites-study. Accessed: 2024-02-24. [↗](https://www.theguardian.com/technology/2023/may/)
13. Cao, Y., Zhou, L., Lee, S., Cabello, L., Chen, M., and Hershcovich, D. [Assessing cross-cultural alignment between chatgpt and human societies: An empirical study](https://md.research.elamur.ai/free/en/papers/cao-2023-assessing-cross-cultural-alignment-between/), 2023.
14. Chakraborty, S., Bedi, A. S., Zhu, S., An, B., Manocha, D., and Huang, F. On the possibilities of ai-generated text detection. arXiv preprint arXiv:2304.04736, 2023. — [arXiv:2304.04736](https://arxiv.org/abs/2304.04736) [find](https://scholar.google.com/scholar?q=Chakraborty%2C+S%2E%2C+Bedi%2C+A%2E+S%2E%2C+Zhu%2C+S%2E%2C+An%2C+B%2E%2C+Manocha%2C+D%2E%2C+and+Huang%2C+F%2E+On+the+possibilities+of+ai%2Dgenerated+text+detection%2E+arXiv+preprint+arXiv%3A2304%2E04736%2C+2023%2E+%E2%80%94+%5BarXiv%3A2304%2E04736%5D%28https%3A%2F%2Farxiv)
15. Chen, Y., Kang, H., Zhai, V., Li, L., Singh, R., and Ramakrishnan, B. GPT-Sentinel: Distinguishing Human and ChatGPT Generated Content. ArXiv, abs/2305.07969, 2023. URL [https://api.semanticscholar.org/CorpusID:258686680](https://api.semanticscholar.org/CorpusID:258686680). — [arXiv:2305.07969](https://arxiv.org/abs/2305.07969) [↗](https://api.semanticscholar.org/CorpusID:258686680)
16. Chiang, Y.-L., Chang, L.-P., Hsieh, W.-T., and Chen, W.-C. Natural Language Watermarking Using Semantic Substitution for Chinese Text. In International Workshop on Digital Watermarking, 2003. URL [https://api.semanticscholar.org/CorpusID:40971354](https://api.semanticscholar.org/CorpusID:40971354). [↗](https://api.semanticscholar.org/CorpusID:40971354)
17. Christin, A. What data can do: A typology of mechanisms. International Journal of Communication, 14:20, 2020. [find](https://scholar.google.com/scholar?q=Christin%2C+A%2E+What+data+can+do%3A+A+typology+of+mechanisms%2E+International+Journal+of+Communication%2C+14%3A20%2C+2020%2E)
18. Clark, E., August, T., Serrano, S., Haduong, N., Gururangan, S., and Smith, N. A. All That’s ‘Human’Is Not Gold: Evaluating Human Evaluation of Generated Text. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Na… [find](https://scholar.google.com/scholar?q=Clark%2C+E%2E%2C+August%2C+T%2E%2C+Serrano%2C+S%2E%2C+Haduong%2C+N%2E%2C+Gururangan%2C+S%2E%2C+and+Smith%2C+N%2E+A%2E+All+That%E2%80%99s+%E2%80%98Human%E2%80%99Is+Not+Gold%3A+Evaluating+Human+Evaluation+of+Generated+Text%2E+In+Proceedings+of+the+59th+Annual+Meetin)
19. Crothers, E., Japkowicz, N., and Viktor, H. Machine Generated Text: A Comprehensive Survey of Threat Models and Detection Methods. arXiv preprint arXiv:2210.07321, 2022. [find](https://scholar.google.com/scholar?q=Crothers%2C+E%2E%2C+Japkowicz%2C+N%2E%2C+and+Viktor%2C+H%2E+Machine+Generated+Text%3A+A+Comprehensive+Survey+of+Threat+Models+and+Detection+Methods%2E+arXiv+preprint+arXiv%3A2210%2E07321%2C+2022%2E)
20. Dycke, N., Kuznetsov, I., and Gurevych, I. Yes-yes-yes: Proactive data collection for ACL rolling review and beyond. In Goldberg, Y., Kozareva, Z., and Zhang, Y. (eds.), Findings of the Association for Computational Linguistics: EMNLP 2022, pp. 300–318, Abu Dhabi, United Arab Emirates, December 2022… [find](https://scholar.google.com/scholar?q=Dycke%2C+N%2E%2C+Kuznetsov%2C+I%2E%2C+and+Gurevych%2C+I%2E+Yes%2Dyes%2Dyes%3A+Proactive+data+collection+for+ACL+rolling+review+and+beyond%2E+In+Goldberg%2C+Y%2E%2C+Kozareva%2C+Z%2E%2C+and+Zhang%2C+Y%2E+%28eds%2E%29%2C+Findings+of+the+Association+fo)
21. Dycke, N., Kuznetsov, I., and Gurevych, I. [NLPeer: A unified resource for the computational study of peer review](https://md.research.elamur.ai/free/en/papers/dycke-2023-nlpeer-unified-resource-computational-study/). In Rogers, A., Boyd-Graber, J., and Okazaki, N. (eds.), Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pp. 5049–5073, To…
22. Else, H. Abstracts written by ChatGPT fool scientists. Nature, Jan 2023. URL [https://www.nature.com/articles/d41586-023-00056-7](https://www.nature.com/articles/d41586-023-00056-7). [↗](https://www.nature.com/articles/d41586-023-00056-7)
23. Fagni, T., Falchi, F., Gambini, M., Martella, A., and Tesconi, M. [TweepFake: About detecting deepfake tweets](https://md.research.elamur.ai/free/en/papers/fagni-2020-tweepfake-about-detecting-deepfake-tweets/). Plos one, 16(5):e0251415, 2021.
24. Fernandez, P., Chaffin, A., Tit, K., Chappelier, V., and Furon, T. Three Bricks to Consolidate Watermarks for Large Language Models. 2023 IEEE International Workshop on Information Forensics and Security (WIFS), pp. 1– 6, 2023. URL [https://api.semanticscholar.org/CorpusID:260351507](https://api.semanticscholar.org/CorpusID:260351507). [↗](https://api.semanticscholar.org/CorpusID:260351507)
25. Fu, Y., Xiong, D., and Dong, Y. Watermarking Conditional Text Generation for AI Detection: Unveiling Challenges and a Semantic-Aware Watermark Remedy. ArXiv, abs/2307.13808, 2023. URL [https://api.semanticscholar.org/CorpusID:](https://api.semanticscholar.org/CorpusID:) 260164516. — [arXiv:2307.13808](https://arxiv.org/abs/2307.13808) [↗](https://api.semanticscholar.org/CorpusID:)
26. Gao, C. A., Howard, F. M., Markov, N. S., Dyer, E. C., Ramesh, S., Luo, Y., and Pearson, A. T. Comparing scientific abstracts generated by ChatGPT to original abstracts using an artificial intelligence output detector, plagiarism detector, and blinded human reviewers. bioRxiv, pp. 2022–12, 2022. [find](https://scholar.google.com/scholar?q=Gao%2C+C%2E+A%2E%2C+Howard%2C+F%2E+M%2E%2C+Markov%2C+N%2E+S%2E%2C+Dyer%2C+E%2E+C%2E%2C+Ramesh%2C+S%2E%2C+Luo%2C+Y%2E%2C+and+Pearson%2C+A%2E+T%2E+Comparing+scientific+abstracts+generated+by+ChatGPT+to+original+abstracts+using+an+artificial+intelligenc)
27. Gehrmann, S., Strobelt, H., and Rush, A. M. [GLTR: Statistical Detection and Visualization of Generated Text](https://md.research.elamur.ai/free/en/papers/gehrmann-2019-gltr-statistical-detection-visualization-generated/). In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics: System Demonstrations, pp. 111–116, 2019.
28. Heikkilä, M. How to spot ai-generated text. MIT Technology Review, Dec 2022. URL [https://www.technologyreview.com/2022/12/19/1065596/](https://www.technologyreview.com/2022/12/19/1065596/) how-to-spot-ai-generated-text/. [↗](https://www.technologyreview.com/2022/12/19/1065596/)
29. Hofmann, V., Kalluri, P. R., Jurafsky, D., and King, S. Dialect prejudice predicts AI decisions about people’s character, employability, and criminality, 2024. — [Dialect prejudice predicts AI decisions about people's character, employability, and criminality](https://md.research.elamur.ai/free/en/papers/hofmann-2024-dialect-prejudice-predicts-ai-decisions/)
30. Hou, A. B., Zhang, J., He, T., Wang, Y., Chuang, Y.-S., Wang, H., Shen, L., Durme, B. V., Khashabi, D., and Tsvetkov, Y. SemStamp: A Semantic Watermark with Paraphrastic Robustness for Text Generation. ArXiv, abs/2310.03991, 2023. URL [https://api.semanticscholar.org/CorpusID:](https://api.semanticscholar.org/CorpusID:) 263831179. [↗](https://api.semanticscholar.org/CorpusID:)
31. Hu, X., Chen, P.-Y., and Ho, T.-Y. [RADAR: Robust AI-Text Detection via Adversarial Learning](https://md.research.elamur.ai/free/en/papers/hu-2023-radar-robust-ai-text-detection/). ArXiv, abs/2307.03838, 2023a. URL [https://api.semanticscholar.org/CorpusID:](https://api.semanticscholar.org/CorpusID:) 259501842. [↗](https://api.semanticscholar.org/CorpusID:)
32. Hu, Z., Chen, L., Wu, X., Wu, Y., Zhang, H., and Huang, H. Unbiased Watermark for Large Language Models. ArXiv, abs/2310.10669, 2023b. URL [https://api.semanticscholar.org/CorpusID:264172471](https://api.semanticscholar.org/CorpusID:264172471). — [arXiv:2310.10669](https://arxiv.org/abs/2310.10669) [↗](https://api.semanticscholar.org/CorpusID:264172471)
33. Ippolito, D., Duckworth, D., Callison-Burch, C., and Eck, D. [Automatic detection of generated text is easiest when humans are fooled](https://md.research.elamur.ai/free/en/papers/ippolito-2019-automatic-detection-generated-text-easiest/). arXiv preprint arXiv:1911.00650, 2019.
34. Jawahar, G., Abdul-Mageed, M., and Lakshmanan, L. V. [Automatic detection of machine generated text: A critical survey](https://md.research.elamur.ai/free/en/papers/jawahar-2020-automatic-detection-machine-generated-text/). arXiv preprint arXiv:2011.01314, 2020.
35. Kelly, S. M. ChatGPT creator pulls AI detection tool due to ‘low rate of accuracy’. CNN Business, Jul 2023. URL [https://www.cnn.com/2023/07/25/tech/openai-ai-detection-tool/index.html](https://www.cnn.com/2023/07/25/tech/openai-ai-detection-tool/index.html). [↗](https://www.cnn.com/2023/07/25/tech/openai-ai-detection-tool/index.html)
36. Kirchenbauer, J., Geiping, J., Wen, Y., Katz, J., Miers, I., and Goldstein, T. [A watermark for large language models](https://md.research.elamur.ai/free/en/papers/kirchenbauer-2023-watermark-large-language-models/). International Conference on Machine Learning, 2023.
37. Kirchner, J. H., Ahmad, L., Aaronson, S., and Leike, J. New AI classifier for indicating AI-written text, 2023. OpenAI. [find](https://scholar.google.com/scholar?q=Kirchner%2C+J%2E+H%2E%2C+Ahmad%2C+L%2E%2C+Aaronson%2C+S%2E%2C+and+Leike%2C+J%2E+New+AI+classifier+for+indicating+AI%2Dwritten+text%2C+2023%2E+OpenAI%2E)
38. Kleinberg, J. and Raghavan, M. Algorithmic monoculture and social welfare. Proceedings of the National Academy of Sciences, 118(22):e2018340118, 2021. [find](https://scholar.google.com/scholar?q=Kleinberg%2C+J%2E+and+Raghavan%2C+M%2E+Algorithmic+monoculture+and+social+welfare%2E+Proceedings+of+the+National+Academy+of+Sciences%2C+118%2822%29%3Ae2018340118%2C+2021%2E)
39. Kreps, S., McCain, R., and Brundage, M. All the news that’s fit to fabricate: Ai-generated text as a tool of media misinformation. Journal of Experimental Political Science, 9(1):104–117, 2022. doi: 10.1017/XPS.2020.37. — [doi:10.1017/xps.2020.37](https://doi.org/10.1017/xps.2020.37) [find](https://scholar.google.com/scholar?q=Kreps%2C+S%2E%2C+McCain%2C+R%2E%2C+and+Brundage%2C+M%2E+All+the+news+that%E2%80%99s+fit+to+fabricate%3A+Ai%2Dgenerated+text+as+a+tool+of+media+misinformation%2E+Journal+of+Experimental+Political+Science%2C+9%281%29%3A104%E2%80%93117%2C+2022%2E+doi%3A+1)
40. Kuditipudi, R., Thickstun, J., Hashimoto, T., and Liang, P. [Robust Distortion-free Watermarks for Language Models](https://md.research.elamur.ai/free/en/papers/kuditipudi-2023-robust-distortion-free-watermarks-language/). ArXiv, abs/2307.15593, 2023. URL [https://api.semanticscholar.org/CorpusID:260315804](https://api.semanticscholar.org/CorpusID:260315804). [↗](https://api.semanticscholar.org/CorpusID:260315804)
41. Lamont, M. How professors think: Inside the curious world of academic judgment. Harvard University Press, 2009. [find](https://scholar.google.com/scholar?q=Lamont%2C+M%2E+How+professors+think%3A+Inside+the+curious+world+of+academic+judgment%2E+Harvard+University+Press%2C+2009%2E)
42. Lamont, M. Toward a comparative sociology of valuation and evaluation. Annual review of sociology, 38:201–221, 2012. [find](https://scholar.google.com/scholar?q=Lamont%2C+M%2E+Toward+a+comparative+sociology+of+valuation+and+evaluation%2E+Annual+review+of+sociology%2C+38%3A201%E2%80%93221%2C+2012%2E)
43. Lavergne, T., Urvoy, T., and Yvon, F. Detecting Fake Content with Relative Entropy Scoring. In Pan, 2008. URL [https://api.semanticscholar.org/CorpusID:12098535](https://api.semanticscholar.org/CorpusID:12098535). [↗](https://api.semanticscholar.org/CorpusID:12098535)
44. Li, Y., Li, Q., Cui, L., Bi, W., Wang, L., Yang, L., Shi, S., and Zhang, Y. Deepfake Text Detection in the Wild. ArXiv, abs/2305.13242, 2023. URL [https://api.semanticscholar.org/CorpusID:258832454](https://api.semanticscholar.org/CorpusID:258832454). — [arXiv:2305.13242](https://arxiv.org/abs/2305.13242) [↗](https://api.semanticscholar.org/CorpusID:258832454)
45. Liang, W., Yuksekgonul, M., Mao, Y., Wu, E., and Zou, J. Y. [GPT detectors are biased against non-native English writers](https://md.research.elamur.ai/free/en/papers/liang-2023-gpt-detectors-biased-against-non/). ArXiv, abs/2304.02819, 2023a.
46. Liang, W., Zhang, Y., Cao, H., Wang, B., Ding, D., Yang, X., Vodrahalli, K., He, S., Smith, D., Yin, Y., McFarland, D., and Zou, J. [Can large language models provide useful feedback on research papers? A large-scale empirical analysis](https://md.research.elamur.ai/free/en/papers/liang-2023-large-language-models-provide/). In arXiv preprint arXiv:2310.01783, 2023b.
47. Liang, W., Zhang, Y., Cao, H., Wang, B., Ding, D., Yang, X., Vodrahalli, K., He, S., Smith, D., Yin, Y., et al. [Can large language models provide useful feedback on research papers? A large-scale empirical analysis](https://md.research.elamur.ai/free/en/papers/liang-2023-large-language-models-provide/). arXiv preprint arXiv:2310.01783, 2023c.
48. Lin, B. Y., Ravichander, A., Lu, X., Dziri, N., Sclar, M., Chandu, K., Bhagavatula, C., and Choi, Y. [The unlocking spell on base llms: Rethinking alignment via in-context learning](https://md.research.elamur.ai/free/en/papers/lin-2023-unlocking-spell-base-llms-rethinking/). arXiv preprint arXiv:2312.01552, 2023.
49. Liu, A., Pan, L., Hu, X., Meng, S., and Wen, L. A Semantic Invariant Robust Watermark for Large Language Models. ArXiv, abs/2310.06356, 2023. URL [https://api.semanticscholar.org/CorpusID:263830310](https://api.semanticscholar.org/CorpusID:263830310). — [arXiv:2310.06356](https://arxiv.org/abs/2310.06356) [↗](https://api.semanticscholar.org/CorpusID:263830310)
50. Liu, Q., Zhou, Y., Huang, J., and Li, G. When chatgpt is gone: Creativity reverts and homogeneity persists, 2024. [find](https://scholar.google.com/scholar?q=Liu%2C+Q%2E%2C+Zhou%2C+Y%2E%2C+Huang%2C+J%2E%2C+and+Li%2C+G%2E+When+chatgpt+is+gone%3A+Creativity+reverts+and+homogeneity+persists%2C+2024%2E)
51. Liu, R. and Shah, N. B. [Reviewergpt? an exploratory study on using large language models for paper reviewing](https://md.research.elamur.ai/free/en/papers/liu-2023-reviewergpt-exploratory-study-using/). arXiv preprint arXiv:2306.00622, 2023.
52. Liu, X., Zhang, Z., Wang, Y., Lan, Y., and Shen, C. CoCo: Coherence-Enhanced Machine-Generated Text Detection Under Data Limitation With Contrastive Learning. ArXiv, abs/2212.10341, 2022. URL [https://api.semanticscholar.org/CorpusID:254877728](https://api.semanticscholar.org/CorpusID:254877728). — [arXiv:2212.10341](https://arxiv.org/abs/2212.10341) [↗](https://api.semanticscholar.org/CorpusID:254877728)
53. Liu, Y., Ott, M., Goyal, N., Du, J., Joshi, M., Chen, D., Levy, O., Lewis, M., Zettlemoyer, L., and Stoyanov, V. [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://md.research.elamur.ai/free/en/papers/liu-2019-roberta-robustly-optimized-bert-pretraining/). ArXiv, abs/1907.11692, 2019.
54. Longino, H. E. Science as social knowledge: Values and objectivity in scientific inquiry. Princeton university press, 1990. [find](https://scholar.google.com/scholar?q=Longino%2C+H%2E+E%2E+Science+as+social+knowledge%3A+Values+and+objectivity+in+scientific+inquiry%2E+Princeton+university+press%2C+1990%2E)
55. Messeri, L. and Crockett, M. J. Artificial intelligence and illusions ofunderstanding in scientific research. Nature, 627:49–58, 2024. [find](https://scholar.google.com/scholar?q=Messeri%2C+L%2E+and+Crockett%2C+M%2E+J%2E+Artificial+intelligence+and+illusions+ofunderstanding+in+scientific+research%2E+Nature%2C+627%3A49%E2%80%9358%2C+2024%2E)
56. Mitchell, E., Lee, Y., Khazatsky, A., Manning, C. D., and Finn, C. [DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature](https://md.research.elamur.ai/free/en/papers/mitchell-2023-detectgpt-zero-shot-machine-generated/). ArXiv, abs/2301.11305, 2023a.
57. Mitchell, E., Lee, Y., Khazatsky, A., Manning, C. D., and Finn, C. [DetectGPT: Zero-shot machine-generated text detection using probability curvature](https://md.research.elamur.ai/free/en/papers/mitchell-2023-detectgpt-zero-shot-machine-generated/). arXiv preprint arXiv:2301.11305, 2023b.
58. Naous, T., Ryan, M. J., Ritter, A., and Xu, W. [Having beer after prayer? measuring cultural bias in large language models](https://md.research.elamur.ai/free/en/papers/naous-2024-having-beer-after-prayer-measuring/), 2024.
59. Newman, D. J. The double dixie cup problem. The American Mathematical Monthly, 67(1):58–61, 1960. [find](https://scholar.google.com/scholar?q=Newman%2C+D%2E+J%2E+The+double+dixie+cup+problem%2E+The+American+Mathematical+Monthly%2C+67%281%29%3A58%E2%80%9361%2C+1960%2E)
60. NewsGuard. Tracking AI-enabled Misinformation: 713 ‘Unreliable AI-Generated News’ Websites (and Counting), Plus the Top False Narratives Generated by Artificial Intelligence Tools, 2023. URL [https://www.newsguardtech.com/special-reports/ai-tracking-center/](https://www.newsguardtech.com/special-reports/ai-tracking-center/). Accessed: 2024-02-24. [↗](https://www.newsguardtech.com/special-reports/ai-tracking-center/)
61. OpenAI. GPT-2: 1.5B release. [https://openai.com/research/gpt-2-1-5b-release](https://openai.com/research/gpt-2-1-5b-release), 2019. Accessed: 2019-11-05. [↗](https://openai.com/research/gpt-2-1-5b-release)
62. Papadimitriou, I., Lopez, K., and Jurafsky, D. Multilingual bert has an accent: Evaluating english influences on fluency in multilingual models, 2023. [find](https://scholar.google.com/scholar?q=Papadimitriou%2C+I%2E%2C+Lopez%2C+K%2E%2C+and+Jurafsky%2C+D%2E+Multilingual+bert+has+an+accent%3A+Evaluating+english+influences+on+fluency+in+multilingual+models%2C+2023%2E)
63. Raffel, C., Shazeer, N., Roberts, A., Lee, K., Narang, S., Matena, M., Zhou, Y., Li, W., and Liu, P. J. [Exploring the limits of transfer learning with a unified text-to-text transformer](https://md.research.elamur.ai/free/en/papers/raffel-2019-exploring-limits-transfer-learning-unified/). The Journal of Machine Learning Research, 21(1):5485–5551, 2020.
64. Sadasivan, V. S., Kumar, A., Balasubramanian, S., Wang, W., and Feizi, S. [Can AI-Generated Text be Reliably Detected?](https://md.research.elamur.ai/free/en/papers/sadasivan-2023-ai-generated-text-reliably-detected/) ArXiv, abs/2303.11156, 2023.
65. Shi, Z., Wang, Y., Yin, F., Chen, X., Chang, K.-W., and Hsieh, C.-J. [Red Teaming Language Model Detectors with Language Models](https://md.research.elamur.ai/free/en/papers/shi-2024-red-teaming-language-model-detectors/). ArXiv, abs/2305.19713, 2023. URL [https://api.semanticscholar.org/CorpusID:258987266](https://api.semanticscholar.org/CorpusID:258987266). [↗](https://api.semanticscholar.org/CorpusID:258987266)
66. Singh, K. and Zou, J. New Evaluation Metrics Capture Quality Degradation due to LLM Watermarking. arXiv preprint arXiv:2312.02382, 2023. — [arXiv:2312.02382](https://arxiv.org/abs/2312.02382) [find](https://scholar.google.com/scholar?q=Singh%2C+K%2E+and+Zou%2C+J%2E+New+Evaluation+Metrics+Capture+Quality+Degradation+due+to+LLM+Watermarking%2E+arXiv+preprint+arXiv%3A2312%2E02382%2C+2023%2E+%E2%80%94+%5BarXiv%3A2312%2E02382%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2312%2E02382%29)
67. Solaiman, I., Brundage, M., Clark, J., Askell, A., Herbert-Voss, A., Wu, J., Radford, A., Krueger, G., Kim, J. W., Kreps, S., et al. [Release strategies and the social impacts of language models](https://md.research.elamur.ai/free/en/papers/solaiman-2019-release-strategies-social-impacts-language/). arXiv preprint arXiv:1908.09203, 2019a.
68. Solaiman, I., Brundage, M., Clark, J., Askell, A., Herbert-Voss, A., Wu, J., Radford, A., and Wang, J. [Release Strategies and the Social Impacts of Language Models](https://md.research.elamur.ai/free/en/papers/solaiman-2019-release-strategies-social-impacts-language/). ArXiv, abs/1908.09203, 2019b. URL [https://api.semanticscholar.org/CorpusID:201666234](https://api.semanticscholar.org/CorpusID:201666234). [↗](https://api.semanticscholar.org/CorpusID:201666234)
69. Sulik, J., Rim, N., Pontikes, E., Evans, J., and Lupyan, G. Why do scientists disagree? PsyArXiv, 2023. Preprint. [find](https://scholar.google.com/scholar?q=Sulik%2C+J%2E%2C+Rim%2C+N%2E%2C+Pontikes%2C+E%2E%2C+Evans%2C+J%2E%2C+and+Lupyan%2C+G%2E+Why+do+scientists+disagree%3F+PsyArXiv%2C+2023%2E+Preprint%2E)
70. Teplitskiy, M., Acuna, D., Elamrani-Raoult, A., Körding, K., and Evans, J. The sociology of scientific validity: How professional networks shape judgement in peer review. Research Policy, 47(9):1825–1841, 2018. [find](https://scholar.google.com/scholar?q=Teplitskiy%2C+M%2E%2C+Acuna%2C+D%2E%2C+Elamrani%2DRaoult%2C+A%2E%2C+Ko%CC%88rding%2C+K%2E%2C+and+Evans%2C+J%2E+The+sociology+of+scientific+validity%3A+How+professional+networks+shape+judgement+in+peer+review%2E+Research+Policy%2C+47%289%29%3A1825%E2%80%93)
71. Topkara, M., Riccardi, G., Hakkani-Tür, D. Z., and Atallah, M. J. Natural language watermarking: challenges in building a practical system. In Electronic imaging, 2006a. URL [https://api.semanticscholar.org/CorpusID:16650373](https://api.semanticscholar.org/CorpusID:16650373). [↗](https://api.semanticscholar.org/CorpusID:16650373)
72. Topkara, U., Topkara, M., and Atallah, M. J. The hiding virtues of ambiguity: quantifiably resilient watermarking of natural language text through synonym substitutions. In Workshop on Multimedia & Security, 2006b. [find](https://scholar.google.com/scholar?q=Topkara%2C+U%2E%2C+Topkara%2C+M%2E%2C+and+Atallah%2C+M%2E+J%2E+The+hiding+virtues+of+ambiguity%3A+quantifiably+resilient+watermarking+of+natural+language+text+through+synonym+substitutions%2E+In+Workshop+on+Multimedia+%26+Se)
73. Tulchinskii, E., Kuznetsov, K., Kushnareva, L., Cherniavskii, D., Barannikov, S., Piontkovskaya, I., Nikolenko, S. I., and Burnaev, E. [Intrinsic Dimension Estimation for Robust Detection of AI-Generated Texts](https://md.research.elamur.ai/free/en/papers/tulchinskii-2023-intrinsic-dimension-estimation-robust-detection/). ArXiv, abs/2306.04723, 2023. URL [https://api.semanticscholar.org/CorpusID:](https://api.semanticscholar.org/CorpusID:) 259108779. [↗](https://api.semanticscholar.org/CorpusID:)
74. Uchendu, A., Le, T., Shu, K., and Lee, D. [Authorship Attribution for Neural Text Generation](https://md.research.elamur.ai/free/en/papers/uchendu-2020-authorship-attribution-neural-text-generation/). In Conference on Empirical Methods in Natural Language Processing, 2020. URL [https://api.semanticscholar.org/CorpusID:221835708](https://api.semanticscholar.org/CorpusID:221835708). [↗](https://api.semanticscholar.org/CorpusID:221835708)
75. Van Noorden, R. and Perkel, J. M. Ai and science: what 1,600 researchers think. Nature, 621(7980):672–675, 2023. [find](https://scholar.google.com/scholar?q=Van+Noorden%2C+R%2E+and+Perkel%2C+J%2E+M%2E+Ai+and+science%3A+what+1%2C600+researchers+think%2E+Nature%2C+621%287980%29%3A672%E2%80%93675%2C+2023%2E)
76. Van Rossum, D. Generative AI Top 150: The World’s Most Used AI Tools. [https://www.flexos.work/learn/generative-ai-top-150](https://www.flexos.work/learn/generative-ai-top-150), February 2024. [↗](https://www.flexos.work/learn/generative-ai-top-150)
77. Walters, W. H. and Wilder, E. I. Fabrication and errors in the bibliographic citations generated by ChatGPT. Scientific Reports, 13(1):14045, 2023. [find](https://scholar.google.com/scholar?q=Walters%2C+W%2E+H%2E+and+Wilder%2C+E%2E+I%2E+Fabrication+and+errors+in+the+bibliographic+citations+generated+by+ChatGPT%2E+Scientific+Reports%2C+13%281%29%3A14045%2C+2023%2E)
78. Weber-Wulff, D., Anohina-Naumeca, A., Bjelobaba, S., Foltýnek, T., Guerrero-Dib, J., Popoola, O., Šigut, P., and Waddington, L. Testing of detection tools for AI-generated text. International Journal for Educational Integrity, 19(1):26, 2023. ISSN 1833-2595. doi: 10.1007/s40979-023-00146-z. URL ht… — [doi:10.1007/s40979-023-00146-z](https://doi.org/10.1007/s40979-023-00146-z) [find](https://scholar.google.com/scholar?q=Weber%2DWulff%2C+D%2E%2C+Anohina%2DNaumeca%2C+A%2E%2C+Bjelobaba%2C+S%2E%2C+Folty%CC%81nek%2C+T%2E%2C+Guerrero%2DDib%2C+J%2E%2C+Popoola%2C+O%2E%2C+S%CC%8Cigut%2C+P%2E%2C+and+Waddington%2C+L%2E+Testing+of+detection+tools+for+AI%2Dgenerated+text%2E+International+Journa)
79. Wolff, M. Attacking Neural Text Detectors. abs/2002.11768, 2020. — [arXiv:2002.11768](https://arxiv.org/abs/2002.11768) [find](https://scholar.google.com/scholar?q=Wolff%2C+M%2E+Attacking+Neural+Text+Detectors%2E+abs%2F2002%2E11768%2C+2020%2E+%E2%80%94+%5BarXiv%3A2002%2E11768%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2002%2E11768%29)
80. ArXiv, Wu, Y., Hu, Z., Zhang, H., and Huang, H. DiPmark: A Stealthy, Efficient and Resilient Watermark for Large Language Models. ArXiv, abs/2310.07710, 2023. URL [https://api.semanticscholar.org/CorpusID:263834753](https://api.semanticscholar.org/CorpusID:263834753). — [arXiv:2310.07710](https://arxiv.org/abs/2310.07710) [↗](https://api.semanticscholar.org/CorpusID:263834753)
81. Yang, X., Cheng, W., Petzold, L., Wang, W. Y., and Chen, H. DNA-GPT: Divergent N-Gram Analysis for Training-Free Detection of GPT-Generated Text. ArXiv, abs/2305.17359, 2023a. URL [https://api.semanticscholar.org/CorpusID:](https://api.semanticscholar.org/CorpusID:) 258960101. — [arXiv:2305.17359](https://arxiv.org/abs/2305.17359) [↗](https://api.semanticscholar.org/CorpusID:)
82. Yang, X., Pan, L., Zhao, X., Chen, H., Petzold, L. R., Wang, W. Y., and Cheng, W. A Survey on Detection of LLMs-Generated Content. ArXiv, abs/2310.15654, 2023b. URL [https://api.semanticscholar.org/CorpusID:264439179](https://api.semanticscholar.org/CorpusID:264439179). [↗](https://api.semanticscholar.org/CorpusID:264439179)
83. Yoo, K., Ahn, W., Jang, J., and Kwak, N. J. Robust Multi-bit Natural Language Watermarking through Invariant Features. In Annual Meeting of the Association for Computational Linguistics, 2023. URL [https://api.semanticscholar.org/CorpusID:259129912](https://api.semanticscholar.org/CorpusID:259129912). [↗](https://api.semanticscholar.org/CorpusID:259129912)
84. Yu, X., Qi, Y., Chen, K., Chen, G., Yang, X., Zhu, P., Zhang, W., and Yu, N. H. GPT Paternity Test: GPT Generated Text Detection with GPT Genetic Inheritance. ArXiv, abs/2305.12519, 2023. URL [https://api.semanticscholar.org/CorpusID:258833423](https://api.semanticscholar.org/CorpusID:258833423). — [arXiv:2305.12519](https://arxiv.org/abs/2305.12519) [↗](https://api.semanticscholar.org/CorpusID:258833423)
85. Zellers, R., Holtzman, A., Rashkin, H., Bisk, Y., Farhadi, A., Roesner, F., and Choi, Y. [Defending Against Neural Fake News](https://md.research.elamur.ai/free/en/papers/zellers-2019-defending-against-neural-fake-news/). ArXiv, abs/1905.12616, 2019. URL [https://api.semanticscholar.org/CorpusID:168169824](https://api.semanticscholar.org/CorpusID:168169824). [↗](https://api.semanticscholar.org/CorpusID:168169824)
86. Zhang, Y.-F., Zhang, Z., Wang, L., Tan, T.-P., and Jin, R. Assaying on the Robustness of Zero-Shot Machine-Generated Text Detectors. ArXiv, abs/2312.12918, 2023. URL [https://api.semanticscholar.org/CorpusID:266375086](https://api.semanticscholar.org/CorpusID:266375086). — [arXiv:2312.12918](https://arxiv.org/abs/2312.12918) [↗](https://api.semanticscholar.org/CorpusID:266375086)
87. Zhao, X., Ananth, P. V., Li, L., and Wang, Y.-X. [Provable Robust Watermarking for AI-Generated Text](https://md.research.elamur.ai/free/en/papers/zhao-2023-provable-robust-watermarking-ai-generated/). ArXiv, abs/2306.17439, 2023. URL [https://api.semanticscholar.org/CorpusID:](https://api.semanticscholar.org/CorpusID:) 259308864. [↗](https://api.semanticscholar.org/CorpusID:)

*Updated 2 November 2026*